A few years ago, I was in a product meeting at Kinzoo, the company I founded, when someone suggested we add streaks.

You know the ones. The little flame emoji that tracks how many days in a row you’ve opened an app. Snapchat uses them. Kids love them. And they’re extremely effective at driving daily engagement.

Lose a streak and it feels like you’ve lost something. So people open the app, even when they otherwise wouldn’t, just to keep the number alive.

We didn’t add them.

I’ve always had a pretty simple rule for product decisions: a feature should benefit both the user and the platform. Streaks felt too one-sided. The platform gets more engagement. The user gets pressure to keep a mostly meaningless number going.

That wasn’t a trade-off I was comfortable with.

I understood the argument, though. I’ve been in enough product meetings to know how these decisions happen. There usually isn’t some evil genius sitting around trying to figure out how to addict children. There’s a product manager looking at retention. There’s a growth target. Someone has an idea that moves the numbers in the right direction.

Streaks work. And that’s the problem.

This week, 47 US states settled with Meta for up to $17 billion over years of product decisions that followed roughly the same logic. The dollar figure made the headlines. I’m not sure it’s the most important part.

Forget the $17 billion for a minute

Meta made $201 billion in revenue last year. The $17 billion settlement gets paid over ten years.

You can decide for yourself how painful that is for a company the size of Meta, but I don’t think the fine changes much. If a company can make billions from a particular business model and eventually pay some fraction of that back in a settlement, the incentive remains pretty obvious.

The product changes are much more interesting.

Daily time limits for teens. Nighttime lockouts. No notifications during school hours. Likes hidden by default. Algorithmic feeds made optional.

Those aren’t features Meta decided to roll out because it suddenly had a change of heart. They are legally mandated changes to some of the most widely used products in the world.

That matters.

Because for years we’ve heard some version of the same defence: these are neutral platforms. People choose how to use them. Parents need to set better limits. Kids need to develop healthier habits.

There’s truth in some of that. Parents absolutely have a role to play, but it ignores the other side of the equation.

Meta employs some of the smartest product designers, engineers and behavioural experts in the world. These products didn’t become extraordinarily good at holding our attention by accident.

Features were tested. Engagement was measured. Retention was studied. The things that worked were kept. And according to the case brought by the states, Meta knew some of those decisions were harmful to young users and made them anyway.

That is the part of this settlement that parents should pay attention to. For a long time, parents worried about these products while being told, explicitly or otherwise, that maybe they were worrying too much. Turns out there was plenty to worry about.

What do you do with that at home?

Probably not much differently tomorrow than you did yesterday. Your teenager still has a phone. Instagram isn’t disappearing. TikTok, YouTube and Snapchat aren’t part of this settlement. And most of the mandated changes will take time to roll out.

But this is a pretty good excuse to talk about it. Not a lecture. A conversation.

“Did you see what happened with Instagram this week? What do you think about it?”

Then see where it goes.

I suspect a lot of teenagers understand the strange relationship they have with these apps better than adults give them credit for. They know when they’re checking something compulsively. They know the weird pressure that comes with likes, streaks and notifications. They know what it feels like when everyone else is somewhere they weren’t invited.

They may just not talk about it very often, and this gives you an opening.

And maybe there’s something useful in being able to tell your kid that some of those things they feel aren’t imaginary. The products really were designed to keep them coming back.

That’s a much more interesting conversation than arguing about screen time.

TL;DR

Some other stories worth your time.

• Kids are better at getting around parental controls than parents think. New UK research found that 91% of parents believe they have control over their kids’ online lives, while 41% of children say they’ve tried to bypass those controls. Half said doing so was easy. Parental controls are useful. They’re just not a substitute for parenting.

• More screen time doesn’t automatically mean worse outcomes. A new study found that children who spent more time on screens didn’t necessarily perform worse on cognitive tests. Some actually performed better. Another reminder that counting minutes tells us a lot less than understanding what kids are actually doing with them.

• Meta would now very much like TikTok and YouTube to have the same rules. One interesting detail buried in the settlement is that some protections depend on competitors adopting similar measures. Meta’s lawyers are already pointing at the other platforms. You can probably guess where regulators look next.

Sean Herman is the founder of Kinzoo, a safe messaging app for kids, and the author of Screen Captured. He writes That Digital Dad to help families navigate this world.

Every few years, the same headline makes the rounds. Big Tech executives, it tells us, seriously limit their own kids’ access to technology. The latest version stars Peter Thiel and Evan Spiegel, who reportedly cap their children’s screen time at 90 minutes a week. But this story is much older than the current news cycle. It first appeared in 2011, when The New York Times profiled a Waldorf school in Los Altos, California. The classrooms were stocked with chalkboards and knitting needles, the campus sat a short drive from Apple’s headquarters, and roughly three-quarters of the students had parents working at companies like Google, Apple and Yahoo.

The story ran again in 2014, when a Times columnist revealed that Steve Jobs never let his own kids use the iPad. Then it ran again when a wave of coverage reported that Bill Gates didn’t allow his children to have phones until they turned 14. It surfaced yet again when we learned that Silicon Valley parents were asking their nannies to sign contracts promising to keep phones away from their kids. And now it’s back for the AI era, with a new cast of characters and the same familiar plot. That adds up to fifteen years. Most of the kids in the current headlines weren’t even born when the first version ran.

It’s easy to see why the story does so well every time it circulates. There’s something tantalizing about the idea that the people building our technology won’t let their own children near it. It suggests they know something the rest of us don’t. They tell us publicly that what they build is safe, but behind closed doors, their behaviour seems to say otherwise.

I won’t speculate on whether it’s all true. There’s enormous pressure on parents to look like they have everything under control, and these particular parents live in the public eye. Maybe their kids get more screen time than the news outlets hear about. Maybe they’re every bit as strict as they claim. I honestly don’t know, and I don’t think it matters much.

What interests me is the story itself. I’m interested in its staying power, and in what that staying power says about what Big Tech has built over the years. I think the story persists because it confirms a distrust that parents already carry. For years, many of us felt that something was off about screen time before we could quite name it. We felt the pull of the feed, the tug of the notification and the strange difficulty of putting certain apps down, and we suspected our kids were feeling it too.

Then the industry kept confirming the suspicion. Former executives came forward to describe products engineered around what one of them called a social validation feedback loop. Leaked internal research showed that companies knew about harms they had downplayed in public. Parents weren’t being paranoid. They were being proven right.

But here’s what I think the story gets subtly wrong. The problem was never technology itself. The problem was a specific set of choices: engagement-maximizing feeds, algorithmic recommendations tuned to keep us scrolling and business models where a child’s attention is the product being sold. Those choices were made by people, which means different people can make different ones.

Notice what the executives’ answer concedes. Rationing screen time down to 90 minutes a week, or withholding a phone until age 14, treats their own products like a vice to be managed. It accepts the premise that the technology can’t be built better, only used less. Coming from the people with the most power to build it better, that’s a striking admission.

And it reinforces the false binary this story sets up every time it circulates: abstinence or addiction. Either you lock the devices away like the executives do, or you surrender your kids to the feed. Those are presented as the only two options.

I don’t accept that they are. The third option is to build technology that’s actually worth sharing with your kids in the first place. That’s what we’re doing at Kinzoo. I didn’t start this company in spite of being a parent. I started it because I’m one, and my own kids aren’t an edge case we design around. They’re our first users.That, to me, is the real litmus test for any technology: are the people who created it excited to share it with the people they love?

If not, that’s certainly a headline.

A deeper dive

Here are a few helpful resources in case you want to really dig into today’s topic:

• Peter Thiel and Evan Spiegel apparently both limit their children to 1.5 hours of screen time a week. This is a great deal less than average. According to the American Academy of Child and Adolescent Psychiatry, children in the U.S. ages 8 to 18 spend 7.5 hours per day watching or using screens. More often than not, when parents hear that tech executives limit their kids to just over 12 minutes a day of screen time, they feel guilty, inadequate or like they’re not doing enough for their own children. But we’re not really comparing apples to apples, since most parents don’t have access to the same resources that tech billionaires do.

• As I mentioned, this idea that tech executives don’t let their own kids near the technology they build is not new. It rears its head every few years. And here’s one of the original articles about Steve Jobs’ approach to tech in his home, which was first published back in 2014.

TL;DR

Too long; didn’t read. It shouldn’t be a full-time job to keep up on industry news, so here is a mercifully quick summary of some other notable developments:

• We’ve all been there: you’re trying to get your kid’s attention, but they’re totally, completely absorbed in their phone. It’s almost a cliché, but kids aren’t the only ones who struggle to put their screens down. According to new research, when parents ignore their kids in favor of their phones, it can exacerbate anxiety and insecurity in children and teens. It’s a good reminder for everyone to put their phones down and enjoy some face-to-face time.

• Instagram recently debuted a controversial new feature that allowed users to create AI-generated imagery of anyone with a public profile. Users who weren’t set to private were automatically opted in, including celebrities and public figures. They had to dig deep into their settings to opt out. After a lot of public backlash, Instagram removed the feature entirely, saying it “missed the mark.” It’s hard to imagine anyone thinking a feature like that would go over well.

And lastly

Here are a few more pieces of original writing from me and my team—just in case you’re keen for more:

• “Screen-free summer” is all the rage on social media, and it seems like an especially tempting goal after reading that some tech executives limit their own kids’ screen time to 1.5 hours a week. If you’re aiming to reduce your kids’ screen time this summer, here are some realistic strategies that avoid the “all or nothing” trap.

• The hit Roblox game Grow a Garden is back with a sequel, and this one introduces a stealing mechanism. Here’s everything parents need to know.

In case you missed it, Snapchat recently rolled out another round of teen safety changes. A few days later, Meta did the same. Read the announcements on their own terms, and they are hard to argue with. More privacy, age-appropriate content, more robust age checks. These are reasonable features, and the sort of thing any parent would have asked for on day one.

But if these protections are so sensible, why did they only arrive in June 2026? It’s not that these Big Tech companies suddenly care deeply about child safety online. It’s that they’re facing unprecedented global pressure from wide-ranging bans to keep under-16s off their platforms. And there’s a pattern that stretches back, where new safeguards seem to follow a lawsuit, a bill, a settlement or some external pressure rather than show up when the product was first built. The features are nice to have, but the timing betrays something fundamental about these companies: they aren’t motivated by child safety. They’re motivated by profit above all else. It’s a pattern we see again and again.

Consider YouTube in 2019. For years, the company told advertisers it was a top destination for kids while telling regulators it had no users under thirteen. That contradiction held until the Federal Trade Commission and the New York Attorney General intervened. On September 4, 2019, Google and YouTube agreed to pay $170 million to settle allegations that they had illegally collected children’s data, the largest penalty ever levied under the Children’s Online Privacy Protection Act. Only as a condition of that settlement did the now-familiar machinery appear: the “made for kids” labels, the limits on data collection, the system requiring creators to flag content aimed at children. But Google didn’t make these changes on its own. The company was forced into it by the FTC.

And then there’s Instagram in the autumn of 2024. On September 17, Meta introduced a sweeping set of default protections for users under eighteen, including private profiles, restricted messaging and parental oversight. The changes were substantial. They also landed one day before the House Energy and Commerce Committee was scheduled to mark up the Kids Online Safety Act. The timing was not lost on the bill’s co-author. Senator Marsha Blackburn called the announcement proof that the industry had no real interest in protecting children, noting that, just like clockwork, the legislation moved forward and the platforms produced a fresh set of guidelines to enforce on themselves.

And then there’s the nudity filter Meta added to teen direct messages in 2024, a tool that automatically blurs explicit images. This is a reasonable feature, plainly. What makes it instructive is a document surfaced through litigation. According to a court filing reported by TechCrunch, the head of Instagram had flagged this very risk in an internal email back in 2018. The awareness was there. The technology was within reach. The product arrived roughly six years later, once the legal and legislative pressure had converged. Knowing was not the thing that produced the fix.

What makes these companies act isn’t some internal guiding principle to keep children safe. If that were true, Instagram would have acted on its nudity filter much faster. But we tend to see action after a federal fine or ahead of a congressional markup. Or, as we’ve seen recently, when momentum builds for far-reaching bans.

The specifics differ each time, but the shape does not. The protective feature trails the pressure. Which brings us back to this year. On March 25, 2026, a Los Angeles jury found Meta and YouTube negligent for designing platforms that harmed young users, the first time a jury had held these companies liable for that harm. During that trial, Mark Zuckerberg testified about various safety features and age checks, at one point stating, “I always wish we would have gotten there sooner, but I think we’re in a better place.” But when these safety features roll out after the fact, it shows that there isn’t a whole lot of appetite to be proactive.

The honest conclusion is not that these companies are incapable of protecting children. They have the engineers, the research and when they choose, the will. We’re seeing that play out in real time with these new protections announced by Snap and Meta (right after the UK announced its under-16 ban). The truth is that capability has rarely been enough on its own. What has reliably moved these platforms is the consequence, the fine, the verdict or the bill. That is a hard thing to build a childhood around. It suggests the safest place for a young person is not a platform that added protections after losing in court, but a product designed, from its first line of code, on the assumption that the children using it deserved protection all along.

A deeper dive

Here are a few helpful resources in case you want to really dig into today’s topic:

• One of the changes Snap made to protect younger children is giving them a “friends-only” experience for their spotlight snaps. That basically attempts to take the open network and shrink it down to trusted contacts only. The fact is, kids are safer when they’re on a platform that’s closed and controlled, and that’s a big reason we started Kinzoo.

• And as for the updates made to Instagram, there are a few changes meant to make the platform safer for teens. There are new content settings, new age assurance and new alerts for parents who supervise their teens’ accounts. As per usual, these features are a reaction to external pressure, not because the company cares about keeping kids safe.

TL;DR

Too long; didn’t read. It shouldn’t be a full-time job to keep up on industry news, so here is a mercifully quick summary of some other notable developments:

• The House has passed The Kids Internet and Digital Safety Act. If it turns into law, minors’ use of disappearing messages would be restricted. It would also require AI chatbots to disclose that they aren’t human and make age verification for platforms hosting pornographic content mandatory.

• Common Sense Media has released its parents’ guide on AI companions and kids. It’s a pretty thorough explainer on the topic and it outlines the main risks for children, including blurring the lines between real and fake relationships, worsening mental health challenges and reinforcing poor decision-making.

And lastly

Here are a few more pieces of original writing from me and my team—just in case you’re keen for more:

• If you’re curious about the ins and outs of the proposed social media ban for under-16s in the UK, check out this explainer piece from my team.

• Governments around the world are also contemplating bans on AI companions. Here’s the latest on what lawmakers are contemplating.

On June 15, Keir Starmer said the UK is going to stop kids under 16 from using social media. As an announcement, it sounds like exactly the kind of decisive action a lot of us have been waiting for. In practice, what the government is actually attempting is one of the hardest needles to thread in tech policy, and I say that as someone who has spent years building technology for children and watching well-intentioned fixes miss the target. So here’s what was announced, what nobody has worked out yet and what you can do at home in the meantime.

The plan is to bar anyone under 16 from holding accounts on the big platforms. The government named Snapchat, TikTok, YouTube, Instagram, Facebook and X. Messaging apps like WhatsApp and Signal are out of scope, and so is YouTube Kids. The government also wants to block under-16s from livestreaming, stop strangers from contacting children on gaming and livestreaming services and require AI “romantic companion” chatbots to verify users are 18 or older. It came out of a national consultation that nine in ten parents backed.

The first regulations are expected to reach Parliament before Christmas, with the rules coming into force in spring 2027, but the distance between the press conference and a working law is enormous. We know enforcement is meant to land on platforms, not on kids or parents, and that families won’t be penalised if a child finds a way around the rules. What we don’t know is how the age checks will actually function, which is sort of the whole question. And nobody has explained how it deals with the workarounds like VPNs and borrowed logins.

Australia is the cautionary tale here. Three months into its own under-16 ban, the country’s online safety regulator found that roughly 70% of kids who already had accounts still had access to at least one major platform, and in most cases they hadn’t even needed a workaround, because the platforms simply hadn’t identified and removed their accounts in the first place.

The experts are split, and what struck me is that both camps want the same outcome. The National Society for the Prevention of Cruelty to Children welcomed the ambition while warning the age checks have to actually work. The skeptics, for their part, have no love for the platforms. Their worry is whether this particular method does anything. The UN human rights office cautioned that bans are easy to get around and risk pushing kids toward less monitored, riskier corners of the internet, and legal specialists noted that once the rules stretch into livestreaming and chatbots, the whole thing gets much harder to police.

Read across those critics and they keep landing in the same place, and it happens to be the argument I agree with. Whether kids can reach these platforms was never the heart of it. What matters is how the platforms behave once anyone, child or adult, is on them. The UN has called for safety to be designed into the architecture from the start instead of leaving parents to manage the fallout, and UNICEF put it plainly that age limits are no substitute for companies fixing their design and moderation. Raising the age on the front door doesn’t make the room behind it safe. A wall is something a determined kid climbs, usually into worse territory. The slower, harder work is changing what’s actually in the room, and that’s the work that holds up.

Which brings me to the part you can do something about today. No regulation will ever protect your kid the way you can. Children who feel they can talk to a parent about their online life make safer choices, ban or no ban. So aim for a kid who comes to you when something online scares them, rather than one who hides it because he’s afraid you’ll take the phone away. Beyond that, get specific about what your kids are actually using and why, because a group chat with their closer friends and a two-hour doomscroll are not the same thing even though they happen on the same device. Some products are engineered to keep kids on the app as long as possible. Others are built to connect them with people who love them, inside boundaries you can see, no strangers, no endless feed, nothing optimising for outrage. Those choices already exist, and you don’t need an act of Parliament to make them.

The announcement does signal that governments may finally be done treating children as an afterthought in how these products get built, and I’m glad for that. But a law arriving in 2027 won’t raise your kid in 2026. That job was always ours, and it comes down to staying close, paying attention and choosing the technology in your home on purpose.

A deeper dive

Here are a few helpful resources in case you want to really dig into today’s topic:

• One outspoken critic of the ban is the Molly Rose Foundation, a British nonprofit that raises awareness about the harms that social media can pose to young people. In response to Keir Starmer’s announcement, they said, “a social media ban will fail to tackle fundamental product safety risks issues and leaves parents with a false sense of safety. A majority of children will continue to use high risk sites that will have no incentive to implement robust protections.”

• Another group that has mixed feelings about the social media ban? Teenagers. The BBC spoke to young influencers who shared some of the ways that social media had improved their lives, whether that was through income, connection or creativity. While they certainly saw the drawbacks, they maintained that the platforms had been a net positive in their lives.

TL;DR

Too long; didn’t read. It shouldn’t be a full-time job to keep up on industry news, so here is a mercifully quick summary of some other notable developments:

• The big AI platforms all have some sort of guardrails built in. In theory, these prevent the bots from sharing anything violent, illegal or sexually explicit. These guardrails are meant to keep users safe, but they’re easier to circumvent than you might think. According to the Washington Post, simple things like asking the bot to pretend to be a rule-breaking character, using poetry or hand-writing instructions and then snapping a photo will work.

• Apple recently announced new child safety features for iOS under Child Accounts. They include new monitoring features like “Ask to Browse” and a feature that requires kids to get permission to browse the internet or visit specific sites. While parents always had the option to set up child accounts, these new features are an evolution of what was already on offer.

And lastly

Here are a few more pieces of original writing from me and my team—just in case you’re keen for more:

• One of the biggest trends on Roblox right now is a weird one: brain rot games. If you’re confused about what that is, check out this explainer piece from my team, complete with everything you need to know.

• Summer break means a lot more free time, and it can feel even harder than usual to keep a balance with screen time. If you want to find the right balance for your kids, check out this piece about summer screen time limits.

A few members of my team went to Vancouver Web Summit recently, and they came back with a line from a Common Sense Media talk that I’ve been thinking about since. The speaker called the group chat a third space: the new mall, the place kids actually go to hang out now. I think that’s exactly right, and there’s a case for thinking about the kind of screen time differently.

Think about where you spent your time as a kid, in the hours that weren’t home and weren’t school. For me, it was some combination of the food court, a friend’s basement, and a parking lot we had no real reason to be standing in. For you, it might have been the diner or the corner store or the bleachers after everyone else had gone home. Sociologists call this the third place. Not the first place where you live or the second where you work, but the in-between. For kids, this is where you start to figure out who you are without an adult setting the agenda.

Sadly, a lot of those places are gone now, or close to it. Malls are dying, the coffee shop has a time limit and a laptop ban, and the kids who do gather in parks tend to get told to move along. And, the kids growing up today have more supervision and less unstructured time than previous generations. So the “third place” did what any space does when its physical version disappears: it moved online. The group chats on iMessage and WhatsApp and Snapchat have turned into a kind of always-open living room, where the inside jokes and the weekend planning and the late-night venting all happen at once.

This is where it gets complicated for parents. When we see our kids glued to their phones all afternoon, we feel a flicker of unease that we wouldn’t if they’d been at the mall or over at a friend’s house. We hear “I met them online” and brace ourselves, even when what’s being described is the same basic social activity we got from loitering near the Orange Julius for three hours. The discomfort is real, but I’ve come to think it’s often aimed at the wrong target.

Not all screen time is equal. The distinction I wish more of us made is between a closed space and an open network. A group chat is a room with a door. There’s no algorithm deciding what we see or hear, there’s no follower count, no strangers drifting in, and no engagement machinery tuned to keep everyone scrolling well past the point of feeling good about it. It’s a small set of people who already know each other. That’s almost the structural opposite of a public platform built to maximize attention, the kind of place that wasn’t designed for a child’s wellbeing to begin with.

What convinces me most is what older kids are choosing on their own. There’s a genuine movement of young people stepping back from their phones right now: flip-phone summers, deleted apps, a return to vinyl and to seeing friends in person. But when we read headlines about what kids are walking away from, it’s worth looking closely at what they’re actually rejecting, because it’s the feed: the scroll, the metrics, the performance. They aren’t abandoning their group chats. Even the flip-phone kids are still texting. When they pare their digital lives down to what they genuinely want to keep, the conversation with their friends is what survives, which says a lot about which part was feeding them and which part was wearing them down.

None of this means group chats are harmless. They can be genuinely cruel, with the sting of being the one left out, the pile-on, the argument that erupts at 2 AM when everyone should be asleep. I’m not romanticizing them, and I’m not suggesting anyone stop paying attention. But it’s worth being precise about the kind of problem this is. The trouble in a group chat is social, the same human friction that played out at the mall and in the cafeteria and at every sleepover, and it’s the sort of thing a kid can learn to handle with a parent nearby. The trouble with an algorithmic feed is built into the design. When a feed keeps your child up too late and leaves them feeling worse, the system isn’t malfunctioning, it’s doing precisely what it was made to do.

That’s why we can’t treat all screen time as a single undifferentiated threat. Wanting to pull our kids back from the feed is a sound instinct. Cutting off the group chat in the same motion gets it backwards, because that’s the connection worth defending. That’s the third place. That’s the mall.

A deeper dive

Here are a few helpful resources in case you want to really dig into today’s topic:

• Earlier this year, Rolling Stone published a piece about third spaces and the way that people are turning to online communities to connect. Here’s what they had to say: “Audiences want to participate, create, get to know each other and be known. They want shared rituals, not endless feeds. Online spaces create accessibility, but the real stickiness is in the relationships that form.” A lot of this resonates from a design perspective. Kids especially need digital spaces that help them connect to the people who matter most.

• There have been lots of articles lately about digital detoxes and analog options, and it seems like Gen Z is leading the charge. But they aren’t rejecting technology writ large; they’re moving away from social media specifically.

TL;DR

Too long; didn’t read. It shouldn’t be a full-time job to keep up on industry news, so here is a mercifully quick summary of some other notable developments:

• We’ve read lots of articles about the dangers of chatbots for kids, but Florida has become the first state to sue OpenAI over child safety. The lawsuit claims that the platform provided information to school shooters, offered guidance on self-harm and addicted young ‌users. Sam Altman, the company’s CEO, was also personally named in the suit.

• And speaking of lawsuits against Big Tech, Texas is suing Netflix for allegedly spying on children and users. Reuters reports that “Texas said that for years, Netflix has falsely represented to consumers that it did not collect or share user data, when it actually tracked and sold viewers’ habits and preferences to commercial data ⁠brokers and advertising technology companies, making billions of dollars a year.”

And lastly

Here are a few more pieces of original writing from me and my team—just in case you’re keen for more:

• Here’s some good news for parents: screen time guidelines are finally catching up to the real world. Here’s our breakdown of the latest AAP advice.

• Kids love to chat with friends while they game, but adult chat apps can be a little dicey. That’s why we built a safer alternative. Learn all about Jams here.

The New York Times recently published a lengthy piece in which a reporter interviewed tweens over the course of a year about how they used chat companions. When she asked them about whether anyone they knew was “dating” the AI bots they talk to, most of them laughed at her, as if she’d asked whether they were dating their favorite book. One of them, a 15-year-old named Quentin who’d spent hundreds of hours talking to character bots, summed it up: “It’s a game. It’s literally ones and zeros.”

If that level of insight from a teen surprises you, it’s because it isn’t what the headlines have trained us to expect. We’ve all absorbed the same steady drip (kids dating chatbots, falling down manosphere rabbit holes, being rewired by their screens) and braced accordingly. But that’s not what these kids are describing. In that Times piece, and in a few others lately, what comes through is more complicated: kids who often understand the technology they’re using better than we think.

In a recent Salon essay, a writer watched a documentary about manosphere influencers with her 17-year-old son, and he was distinctly unimpressed. The influencers, he noted, “make their money scamming people,” and he couldn’t think of anyone he knew who took them seriously, not even his “stupidest friends.” Again, that’s not the story most of us are carrying around in our heads. And I think the gap between the story we expect and the one these kids are actually telling warrants a closer look.

Here’s the narrative we marinate in as parents: passive kids, predatory platforms, inevitable harm. And, it isn’t baseless. I’ve spent years arguing that some of these products are deliberately designed to exploit the people using them. But the narrative has a blind spot. It almost never includes the kids’ own account of what they’re doing. We watch them watching the screen. We rarely ask about it.

When you do ask, you often find discernment. The teens describing chatbots talk about them as games, as interactive fan fiction, as a way to kill boredom or practice writing. They notice the seams, and several even complained that the bots kept steering conversations toward flirtation they hadn’t asked for, which annoyed them. They are not, by and large, confusing software for a soulmate. The kids parsing the manosphere read those influencers as snake oil salesmen and entertainment, not gurus. In both cases, the young people held a more accurate model of the technology than the anxious adult expected.

But I’d be doing exactly what I criticize if I stopped there, because the same reporting that reassured me also shows the edges. One of Quentin’s friends confided that he’d once spent fourteen hours straight talking to bots and couldn’t pull himself away. Researchers in the piece are blunt that the teens most drawn to chatbots tend to be the lonelier, less socially experienced ones. Those are the kids already in a harder spot, who can get pushed further down. And the genuine tragedies are real. I’ve written before about the dangers. So the honest synthesis isn’t “kids are fine.” It’s that most kids are playing, discernment is real but not universal, and the kids most at risk are precisely the ones least able to narrate their experience cheerfully to a reporter. Agency and vulnerability both exist, they’re just not evenly distributed.

The data backs up the nuance. Pew’s recent survey of teens and AI found that across all teens and all uses, the top activities are information-seeking (57%) and schoolwork (54%), with “for fun” close behind (47%). That’s a very different picture than the wall-to-wall companionship the headlines imply. So what do we do with this? Two things.

First, we talk to our kids. I know this is the least technological advice a tech founder can give. But curiosity beats surveillance every time. The question “what do you actually do on there?” gets you a truer and usually more reassuring picture than any monitoring dashboard, and it keeps the channel open for the day something is wrong.

And second, we respect their intelligence. The kid who feels respected is the kid who tells you when something bad is happening in their online life. Respect is what makes vigilance work.

Most of the adults are anxious, and most of the kids are genuinely fine. Our job isn’t blanket alarm or blanket reassurance. It’s paying close enough attention to tell the difference. And that starts with a conversation, not a dashboard.

A deeper dive

Here are a few helpful resources in case you want to really dig into today’s topic:

• “Wow, this is garbage, but fun.” That’s the way Quentin described chatting with AI chatbot Talkie. In the lengthy article, he says he talked with the bots when he was bored, down or experiencing drama with friends. According to researchers, it’s common for teens to treat chatbots as friends they can talk to about anything. But according to the reporter, many teens have a better grasp on the limitations of these systems than some adults.

• The Salon reporters teenage son explained to her that the videos the influencers produce are entertainment, not instructional videos. “It’s one of the biggest stereotypes there is: the big buff dude is a misogynist,” he explains. “Whether they’re serious or they’re doing it as a bit or a full act, it’s funny no matter what.” A lot of the rhetoric around the manosphere involves a lot of handwringing by experts who worry about young people looking to these videos for life advice. And while I don’t think it’s great as entertainment, that’s a better way to think about it than instructions.

TL;DR

Too long; didn’t read. It shouldn’t be a full-time job to keep up on industry news, so here is a mercifully quick summary of some other notable developments:

• Lawmakers are eyeing a ban on chatbots for kids, similar to the social media bans rolled out in Australia. A bill advanced in the Senate that would compel companies to verify users’ ages and ensure children aren’t using the bots. And, it would make it a crime to “knowingly provide a chatbot that engages minors in sexually explicit behavior or coerces them to die by suicide.” While I think it’s great that lawmakers are paying attention to the platforms kids are using, bans aren’t always the most effective way to make children safer.

• There’s been a lot of ink spilled on AI and its effects on, well, everything. One author recently wrote a book on the effect of artificial intelligence on truth, and, ironically, it’s come out that some of the quotes he cites are AI hallucinations. He admits that he used AI in the research, outlining and editing process, and that’s where the misquotes came from.

And lastly

Here are a few more pieces of original writing from me and my team—just in case you’re keen for more:

• Adopt Me! on Roblox seems like a pretty innocuous game, but it has some safety considerations parents should be aware of. Here’s what parents need to know.

• And if your kids seem a little on edge after spending time on screens, this might be the culprit.

Walk into any school pickup line, scroll through any parenting subreddit, sit in on any cabinet meeting where children’s wellbeing is on the agenda, and you will eventually hear it. The screen time debate is back. It never really left, of course. But for a few years, it got a bit quieter. Now it’s louder than it has been in a decade, and the shape of the conversation has changed in ways worth paying attention to.

For most of the last twenty-five years, parents had one dominant mental model for thinking about technology: the clock. Two hours a day. None for kids under two. The timer on the iPad, the arguments at bedtime, the guilt at the end of a long week. Almost every conversation about kids and screens, formal or informal, traced back to a number.

That number had a specific origin. The American Academy of Pediatrics first introduced a time-bound guideline in a 2001 policy statement on children and television, and reaffirmed it in updates over the following decade, including the American Academy of Pediatrics’ 2016 recommendations. By the time most of today’s parents were raising kids, those rules had calcified into a single mental shorthand: the clock. That meant no digital media for kids under 2. One hour a day for kids 2-5. Consistent limits for everyone older. When I started Kinzoo in 2018, the number one question parents would ask me was “how much screen time?” The clock was doing the work.

Even then, I thought we were measuring the wrong thing. Arguing over heroin and potatoes was never going to get us anywhere. A child spending an hour animating their own video is very different than a child spending an hour scrolling social media. Pretending that those are the same, because they spent the same number of minutes looking at a screen, was always going to fail us.

You remember what the pandemic years were like. We were all home, all the time, with kids who needed to be educated, entertained and connected to family we couldn’t visit. Screens did all of it. They were the classroom, the playground, the babysitter and the family reunion, sometimes in the same afternoon.

The screen time clock didn’t survive the pandemic. It couldn’t. Experts who’d spent careers cautioning against screen time were publishing op-eds essentially saying: don’t beat yourself up, screens aren’t that bad. Pediatricians softened their messaging. Parents retired the timer. For a moment, it felt like the entire screen time debate had been stuffed into a drawer.

But in the years since, parents have started to understand something the clock could never measure: a lot of what kids encounter on screens isn’t neutral. Big tech companies have spent decades refining their platforms to maximize engagement, and they’ve been so successful at it, particularly with young users they were ostensibly never targeting, that some now argue the failure to keep kids off these platforms wasn’t a failure at all. It was a strategy.

Parents have seen this play out at home. The platforms that grab hold and won’t let go. The algorithms that reach for the most extreme version of whatever a kid showed mild interest in. The strangers with friend requests. The pressure cooker of social comparison disguised as a feed.

I wrote about all of this in 2019 when I published my first book. At the time, I felt like I was part of a vocal minority, but now that’s changing. Social media bans for young users are passing in jurisdictions around the world. Landmark lawsuits are showing that tech platforms can be held accountable for design choices that knowingly harmed kids. Whistleblowers have given us internal documents that confirm what parents have suspected all along. And understandably, parents are returning to the question that started this whole conversation. How much screen time is safe?

I get why we still default to the clock. It’s the framework we were handed, and when something feels overwhelming, reaching for a number feels like control. But the answer was never hours and minutes. It was always content, context and design. What is my child doing on this screen? Who built this experience, and what are they trying to get out of it? What is this displacing in my child’s day? Those are harder questions. But they’re the right ones.

In January 2026, the AAP released its newest policy statement, Digital Ecosystems, Children, and Adolescents. The statement effectively retires hour-based screen time limits for most children. In place of the clock, the AAP introduces what it calls the 5 Cs: Child, Content, Calm, Crowding Out, and Communication. Parents are encouraged to think about who their child is, what they’re consuming, whether the experience helps them regulate or dysregulate them, whether screens are crowding out essentials like sleep and movement, and how the family talks about media together.

But the deepest shift is in the framing. The new statement names something many of us have been saying for years: that a child’s relationship with technology isn’t just a family-level problem. It’s shaped by the design choices of platforms, the commercial incentives behind those choices, and the policy environment that allows them. The statement acknowledges, plainly, that digital media is “designed to maximize time, reach and activity on platforms” and is “not necessarily optimized for children’s well-being.”

It moves the question from how do families cope with this to why is the ecosystem built this way in the first place. It puts the responsibility where it belongs. Not entirely on parents, and certainly not on kids, but on the systems and incentives that produced this environment. That’s the conversation we should have been having all along.

A lot of people have been making versions of this argument for years. Researchers, pediatricians, educators, advocates, other parents. The shift is collective and overdue. For twenty-five years, the screen time conversation was anchored to a number that was always going to fail us. We’ve finally started asking better questions. Not how long, but what, with whom, and why. Not how do we limit screens, but how do we build a digital world that’s actually good for kids?

The screen time debate is back. But it’s a different debate now, and a better one. This new conversation gives us a better chance to give our kids the digital world they deserve.

A deeper dive

Here are a few helpful resources in case you want to really dig into today’s topic:

• If you want to see exactly how much the conversation around screen time has changed, dig into the new policy statement from the American Academy of Pediatrics. Today, their advice is much more nuanced, with frameworks and recommendations—and a deeper discussion about what kids are doing on screens, not just how much time they’re spending with them. They move away from time limits, only mentioning them once for families who want that advice.

• In an effort to protect kids from some of the harms of excess screen time, lots of schools have adopted cell phone bans, and they’ve been popular among educators, parents and policy makers. But according to new research, the results from the bans are somewhat mixed. Teachers love them and report that students are less distracted in class. Students also report an increase in overall wellbeing. But the bans did not improve test scores or students’ perception of bullying. As well, some schools reported more suspensions after bans went into effect.

TL;DR

Too long; didn’t read. It shouldn’t be a full-time job to keep up on industry news, so here is a mercifully quick summary of some other notable developments:

• Momentum is building to ban AI chatbots for children. Back in April, the Senate Judiciary Committee voted to advance a bill that would make AI companies implement age verification to stop minors from using their platforms. It would also mandate that AI platforms regularly remind all users that they’re AI, and impose criminal penalties on makers whose platforms have sexual conversations with kids or encourage suicide.

• While gaming platforms weren’t included in Australia’s social media ban, the online safety commissioner is turning her attention on them now. She’s ordered Roblox, Minecraft, Fortnite and Steam to explain how they protect children from grooming, exploitation and extremism. If they don’t respond within 30 days, they’ll face fines of up to $590,783 a day.

And lastly

Here are a few more pieces of original writing from me and my team—just in case you’re keen for more:

• A lot of parents are wondering whether screen time is making their kids anxious. If yours melt down when you take away the iPad, this article is for you.

• And if you want a clear breakdown of the action points in the new APP guidelines, check out this piece from my team.

Imagine this: your kid logs onto a digital platform where they have a profile page and a customizable avatar. They have a friends list. They can chat, video call and even “whisper” to the people on it. Their avatars can hang out together on a digital beach or in a digital forest. Depending on their settings, other users can see their friend counts, who they’re following, and who’s following them. And it’s all so engaging that lots of kids spend hours a day hanging out virtually.

Does that sound like a social network to you? It certainly does to me. I’ve just described Roblox, one of the most popular digital platforms for kids in the world. The company reported 144 million daily active users at the end of 2025, roughly 50 million of whom are under 13. But Roblox isn’t a game. Roblox is a social network, and we need to start treating it like one.

Ask any parent, and they’ll probably tell you: their kids hang out on Roblox. But a lot of parents file the platform mentally under “gaming,” not “social media.” And they’re not alone. Lawmakers crafting social media legislation and age-based restrictions have, so far, largely left Roblox off the list. That classification isn’t neutral, and right now, it’s doing the wrong work.

When parents hear “social media,” they think about risks: strangers, inappropriate content, predators, compulsive use. They set rules, have conversations, check in. When they hear “gaming,” they think about screen time and maybe in-app purchases.

Those are different mental models. They lead to different decisions about supervision, settings and which conversations to have with our kids. A parent who would never let their 10-year-old onto Instagram might hand them Roblox without a second thought, because one is “social” and the other is “just a game.”

But Roblox has a profile, a friends list, DMs, voice chat and follower counts. Its Home page, which the company says drives over 90% of traffic on the platform, is powered by a “Recommended for You” algorithm that serves up user-generated content personalized to each player, drawing on engagement signals, retention and friend activity. That sounds a lot like the basic architecture that defines TikTok and Instagram. The fact that the social layer for Roblox is wrapped around games doesn’t change what the platform does. It just changes how we perceive it.

Lawmakers fall into the same trap. The wave of social media legislation moving through statehouses and parliaments, age verification bills, restrictions on accounts for minors, mandated parental controls, was largely written with Instagram, Snap and TikTok in mind. Roblox often sits outside those frameworks, even though tens of millions of children are using it in ways that look a lot more like social media than like Mario Kart.

While the cultural and legislative classification has lagged, the courts have started to see Roblox more clearly. As of April 2026, there are 146 lawsuits consolidated in a federal multidistrict litigation in the Northern District of California. The cases allege that Roblox facilitated child sexual exploitation and prioritized growth over safety. That number was 85 at the start of the year, meaning it nearly doubled in four months. And more lawsuits are expected.

Individual states have been moving, too. Nevada, Alabama and West Virginia have all recently settled with the company, collecting around $12 million each along with commitments to strengthen child safety features. This isn’t a fringe concern anymore. It’s a coordinated legal reckoning, and the through line in almost all of it is the same: Roblox functioned as a place where adults could reach children, and the safety infrastructure didn’t match that reality.

To its credit, Roblox has made meaningful changes. The company has rolled out new account types with age-based restrictions, mandatory age verification to chat, tighter default settings for its youngest users and a tiered system that will further segment the experience by age group. Those are real improvements, and parents should know about them and use them.

But better safety features on a platform won’t change what the platform is. You can make a social network safer. That doesn’t make it not a social network. And as long as parents and policymakers think of Roblox as “just a game,” the conversations we need to be having about supervision, age-appropriate use and where Roblox fits in the broader ecosystem of platforms competing for our kids’ time, will keep missing the mark.

This isn’t a call to ban Roblox. My kids use it. Most kids use it. That’s not going away, and pretending otherwise isn’t useful. What is useful is to audit the platforms your kids use by function, not by category. If it has friends lists, profiles, DMs and user-generated content from strangers, treat it like a social network. Because functionally, that’s what it is. Have the same conversations. Set the same kinds of limits. Check in with the same kind of attention.

And for lawmakers writing the next round of online safety legislation: don’t let the word “gaming” be a loophole. The social networks kids actually use don’t all look like Instagram. Increasingly, the most influential platforms in a child’s life are the ones we’ve been calling games. Our policy frameworks need to reflect that, or they’ll keep missing the platforms where the most kids actually are. The next generation of social media won’t announce itself as social media. It’ll come dressed as something else: a game, a creative tool, a virtual friend. Our job as parents is to see past the costume and name what’s actually there. Roblox is a social network. Let’s start talking about it like one.

A deeper dive

Here are a few helpful resources in case you want to really dig into today’s topic:

• Here’s what parents can expect from Roblox’s two new age-based accounts: one for users 5-8 years old, and one for users 9-15 years old. Users’ ages will be determined using Roblox’s global age check technology (an AI face scan to determine age) or by a verified parent account. The youngest users will have all communication disabled by default and will only have access to content rated mild or minimal. Default chat settings for users 9-15 won’t change, but they’ll be limited to content up to a moderate rating. This announcement comes after a turbulent time for the company marked by lawsuits and allegations that the platform wasn’t doing enough to protect its youngest users.

• Back when Roblox announced that users would have to undergo AI age estimation, the CEO sat for an interview with Hard Fork. The hosts asked him a lot of difficult questions about predation and child safety, and a generous reading is that he was pretty unprepared. The way he evaded the topic almost made it look like he was in denial that there was even a safety issue. I understand that it’s a tough thing to discuss, but as a parent, you definitely want to see the CEO of the company engage with questions on child safety when there’s been so much negative press about it.

TL;DR

Too long; didn’t read. It shouldn’t be a full-time job to keep up on industry news, so here is a mercifully quick summary of some other notable developments:

• Pew Research just released a new report outlining how teens feel about their social media use. Roughly nine in 10 teens use Snapchat, Instagram and TikTok for entertainment. And beyond that, teens are turning to these platforms to connect with people they know and to find people who share similar interests. Across all three platforms, teens report that they neither hurt nor helped their general mental health.

• Pew Research also asked parents to rate how social media affected their teens’ mental health, and they were decidedly more negative in their views. When asked if these platforms helped or hurt, 24% of parents said hurt, and only 8% said helped.

And lastly

Here are a few more pieces of original writing from me and my team—just in case you’re keen for more:

• One of the most popular games on Roblox right now is a horror-themed game called 99 Nights in the Forest. Here’s a rundown of everything parents need to know.

• Another popular game on the platform is decidedly less spooky, but still has a tendency to suck players in. Here’s what you need to know about Grow a Garden.

Louis Theroux has a new documentary on Netflix, Inside the Manosphere. If you’re a parent and you haven’t watched it yet, it’s worth checking out. But fair warning, it’s not easy viewing.

Theroux shows up, he’s disarmingly polite and he just lets influencers from the manosphere talk. And the men he’s talking to, the podcasters, the lifestyle gurus, the self-proclaimed alphas racking up millions of views telling young men how to be men, they talk a lot. They can’t help themselves. The camera is basically oxygen to them.

But there are moments where the facade cracks. It tends to happen when Theroux gets them next to someone who actually knows them: a mother, a wife, a girlfriend. The volume drops a little. The certainty gets wobblier. The persona, which had seemed so total, suddenly looks like what it is: a performance. One that happens to be generating an enormous amount of money. And these are multi-million dollar operations, complete with courses, coaching programs, memberships and merchandise. The more extreme the content, the wider the audience, the bigger the funnel. The outrage isn’t a byproduct. It is the product.

That’s actually the thing that stuck with me. Not the opinions they share (as ugly as some of them are), but the performance of it all. These guys are always on. Every word is content. Every interaction is a bit. And you start to wonder, watching it: did they start out like this? Or did something pull them here? I think it’s the latter, and that matters more than we’re giving it credit for.

Here’s what I keep coming back to. These don’t seem like men who woke up one day with a fully formed worldview and went looking for an audience. They seem like grifters who found out through trial and error what the algorithm rewarded. Extreme positions travel further than moderate ones. Outrage gets more engagement than nuance. Every time they pushed a little further and the numbers went up, the feedback loop got a little tighter. The platform wasn’t just broadcasting their message. It was helping to shape it, and actively pulling it toward whatever version performed best.

That’s a different problem than a few bad actors with big platforms. That’s a system producing a specific kind of content, at scale, because that’s exactly what it was built to produce. I’ve written before about how the algorithm is optimized for extreme content, and this is just an extreme example of that in action.

And the audience at the end of it (mostly young men, a lot of them teenagers), they’re not idiots. They’re looking for something real. Structure. Identity. Even entertainment. Those are legitimate things to want. The tragedy is that what they’re getting back has been optimized not to answer those questions but to keep them watching long enough to sell them a very dubious course.

I don’t think Theroux quite goes there. He’s more interested in the men than the machine, which makes for a compelling documentary. But it lets the algorithm off the hook, and it’s already working on the next guy.

A deeper dive

Here are a few helpful resources in case you want to really dig into today’s topic:

• For those of you with older children, it can be a worthwhile exercise to watch the documentary with them. One author watched with her 17-year-old son and discovered that, while he was familiar with the influencers in the documentary, he didn’t know anyone who took them seriously. “It’s one of the biggest stereotypes there is: the big buff dude is a misogynist,” her son explains. “Whether they’re serious or they’re doing it as a bit or a full act, it’s funny no matter what.”

• While a lot of people focus on how boys fall into the manosphere, new research out of Australia focuses on how some leave it. According to The Conversation, the disillusionment happens when experiences in the real world don’t match up with what they’re seeing online: “Over time, some participants begin to notice the gap between the ideology promoted in these spaces and their everyday experiences. Positive interactions with women, supportive friendships, or simply observing that relationships in the real world do not follow the rigid rules promoted online can begin to undermine the worldview.”

TL;DR

Too long; didn’t read. It shouldn’t be a full-time job to keep up on industry news, so here is a mercifully quick summary of some other notable developments:

• Now that we’re more than 3 months into Australia’s social media ban, we’re starting to get a clearer picture of how the policy is working. According to some estimates, about two-thirds of children under 16 still have access to their social accounts. Australia’s communication minister thinks that Big Tech companies simply aren’t trying that hard to comply: “What we are seeing is evidence of the absolute bare minimum from social media companies, it’s straight out of the big tech playbook … They obfuscate, they try to throw doubt on any regulation.” Sounds about right to me.

• The Google AI summary has changed the way a lot of us search the internet. And according to new research, it’s accurate about nine times out of ten. While that sounds like a passing grade, it hits different when you realize that it translates into hundreds of thousands of incorrect answers every minute. Another worrying finding? Facebook and Reddit were the second- and fourth-most-cited sources.

And lastly

Here are a few more pieces of original writing from me and my team—just in case you’re keen for more:

• Last month’s verdict against Meta and Google has a lot of parents questioning the apps their kids use every day. If you’re wondering how you can keep your kids safer online, check out this post from my team.

• Google’s AI platform, Gemini, is available to kids under 13. Here’s a post from my team outlining everything you need to know.

Here’s a pattern you’ve probably seen before: A new technology arrives. It promises to revolutionize how we connect, learn and live our lives. Everyone rushes to adopt it. Then, years later, we fall out of love with that technology and start frantically trying to undo the damage.

We did this with social media. We’re watching as Australia and Indonesia enact bans for kids under 16, attempting to block children from the platforms we enthusiastically welcomed into our lives. And while we’re not in the honeymoon phase for Instagram or Snapchat anymore, we’re certainly repeating the same pattern with the latest, greatest technology promising to change our world: artificial intelligence.

There’s actually a name for this cycle. Gartner, a research and consulting firm, calls it the Hype Cycle: enthusiasm, rapid adoption and then disenchantment. Social media and smartphones are deep into the disenchantment phase, hence the lawsuits, the mental health crises and the desperate regulatory scrambles.

AI for kids is just entering the rapid adoption phase, but here’s what’s different: this time, the enthusiasm isn’t even really there. At Kinzoo, we recently surveyed our community of tech-savvy parents and asked how comfortable they were with their children using AI products. Less than 20% said they were comfortable with it. And yet AI is already showing up in schools, apps marketed to children, toys, tutoring software and platforms kids use every day. With social media, we at least believed in the promise. This time, we apparently don’t, and it’s happening anyway.

At the very least, parents want companies to prove their AI products are safe for children before releasing them. They want transparency about how these systems work, what data they’re collecting and what the long-term effects might be. And yet, companies are forging ahead with implementation.

When Australia banned social media for children under 16, it made headlines. When Indonesia followed suit, it became a trend. Bans are what happen when you’ve already lost control. They’re the last resort when all the earlier opportunities to build guardrails, demand transparency and enforce accountability have been squandered. We could have required platforms to prove they were safe for kids before rolling them out, but we didn’t. So now we get blanket bans, because governments and parents feel like that’s all that’s left.

By the time we noticed the harms of social media, the technology was so embedded in our lives that removing it felt impossible. That’s the trap of the hype cycle: we adopt during the enthusiasm phase when everything looks like upside. The problems emerge later, when we’re already locked in economically, socially and behaviorally. With AI, it feels like we’re being rushed past the early phase entirely.

To be clear, I believe AI has the potential to be a positive force for children and youth - if the right incentives are in place. Tools that unlock learning and creativity for kids are genuinely exciting. And the reality is, like every major technology before it, AI will be part of our children’s lives. That means we should be focused on teaching them how to use it responsibly, early on.

But we’re also starting to see familiar patterns emerge. Development is moving fast, often prioritizing engagement over well-being, especially with things like AI companions. And just like with social media, meaningful action tends to come only after something goes wrong and makes headlines.

Breaking the cycle would mean doing something radically different: demanding guardrails before adoption, not after harm. It would mean listening to the more than 80% of parents who say they’re not comfortable. It would mean asking “is this actually good for kids?” before asking “how fast can we scale this?”

But I’m not particularly optimistic that’s going to happen. Not from the big platforms anyway. The incentives are all wrong. There’s money in rapid adoption, and there’s pressure to stay competitive. And so we’ll probably keep doing what we’ve always done: adopt enthusiastically, regret eventually and scramble to protect the next generation from whatever comes next.

I don’t know if we’ll ban AI for kids in ten years the way we’re banning social media now. But I do know this: we’ve never been better positioned to get it right. We have the data from the social media years. We have parents who are paying attention. And we have the awareness that the decisions we make in the next few years will matter enormously. That’s not nothing.

A deeper dive

Here are a few helpful resources in case you want to really dig into today’s topic:

• It’s interesting that we can be at different points in the hype cycle for different technologies. Certainly, the backlash against smartphones and social media has been mounting for years. All while governments are investigating whether AI can help support kids’ learning in the classroom. In 2011, an UNESCO brief argued that “social media can help schools allow the real world into their classrooms, and hence prepare students for a real, better future.” It makes this moment of AI hype seem a little like deja vu.

• Speaking of the classroom, parents are noticing a trend of growing screen time at school. Some students are receiving iPads in kindergarten to play games and watch shows, and the amount and type of screen time is troubling parents. One mother told the New York Times that she was recording a video of her daughter dancing to send to her grandparents, and when the song ended, her daughter looked at the camera and said, “If you like what you saw, click below to subscribe.” Wonder where she got that.

TL;DR

Too long; didn’t read. It shouldn’t be a full-time job to keep up on industry news, so here is a mercifully quick summary of some other notable developments:

• Have you heard of Luddite teens? There’s a growing movement among younger people to move away from technology—or at the very least, critically question how they’re being asked to use it. When one journalist spoke with teens about their tech use, this is what she learned: “Some of these teens find that being extremely online is a depressing way to live, and they want a future that involves more embodied activity (music, animal care and crafting all came up as nonscreen alternatives) and real-life connection. But this desire has to come from them, not as a lecture from the adults around them.”

• WhatsApp has just launched parent-supervised accounts for younger users. While the platform has 13+ terms of service, it’s well-known that there are lots of kids on the app. And, Meta, the company that owns WhatsApp is just now launching tools to help parents manage that. According to Tech Crunch, the new accounts won’t have ads, and by default, parents would receive an alert when pre-teens add, block or report a contact.

And lastly

Here are a few more pieces of original writing from me and my team—just in case you’re keen for more:

• If you’re looking for a messaging app for your family, you might want to consider an alternative to WhatsApp. My team explains why here.

• MrBeast has been in the news a lot lately for his expansion in the world of kids’ finance. If you’re wondering who he is, check out my team’s breakdown here.

Artificial intelligence is being integrated into products at extraordinary speed. For adults, it shows up as AI-powered search, automated writing and smarter tools. For children, it is appearing in homework platforms, creative apps and even teddy bears. The industry often frames this as inevitable progress. But, many parents are not so sure.

We recently surveyed our own community of Kinzoo users and asked how comfortable they were with their children using AI products. These are digitally literate Lifeguard Parents who embrace technology in their own lives and want their children to benefit from it. They are not trying to shield their families from innovation. They believe in staying involved rather than pushing kids into complex systems alone. And, fewer than 20 percent said they were comfortable with AI.

That gap between industry enthusiasm and parental caution is instructive. Companies building for children face a structural tension that extends beyond individual features. Innovation cycles are accelerating. Investors expect momentum. Engagement metrics drive valuation. Distribution partnerships and monetization models are often designed to maximize scale. At the same time, children are uniquely impressionable users. Decisions that increase reach, retention or revenue can also introduce developmental, privacy or safety risks.

At Kinzoo, those tensions are not theoretical. They surface not only in product design discussions, but in broader business decisions. When we evaluate a new feature, consider a distribution strategy or assess a potential growth lever, we start with a practical question: if this goes wrong, who bears the cost? Adults can contextualize confusing or manipulative outputs. Children may not. Adults can step away from persuasive systems. Children may internalize them. That lens changes not just what we build, but how we grow.

There are well-established mechanics that reliably accelerate growth in consumer technology: open networks, viral loops, algorithmic recommendations optimized for retention and targeted advertising. They are powerful because they work. They are also the same mechanisms that have required years of retrofitted safeguards once harm became visible. We have chosen to forgo those mechanics, even when they could produce faster growth or stronger short-term metrics. That choice may limit network effects and slow expansion. It also reduces exposure.

When we’re not sure what to do, we slow down. We revisit our mission and ask whether a proposed decision advances it. We seek additional input from parents, look for emerging research and stress-test worst-case scenarios. Sometimes we shelve an idea entirely. Sometimes we rethink it. But we don’t rush forward just because the market is moving.

This is not a claim to moral superiority. It is an acknowledgment of trade-offs. Building for children means accepting that some opportunities aren’t right. It means deciding in advance what risks are unacceptable rather than recalibrating only after public backlash.

As AI becomes embedded in children’s products, the stakes increase. Generative systems can simulate companionship, respond dynamically to emotional cues and operate in ways that are difficult even for adults to fully interpret. The developmental implications remain uncertain. That uncertainty alone warrants caution.

Transparency does not require publishing every internal debate. It means being clear about the principles that guide our decisions. It means publicly sharing research on parent sentiment, even when it reveals skepticism. It means acknowledging that growth targets and families’ needs can pull in opposite directions, and stating openly which one has priority.

The industry’s history with child safety has not earned unconditional trust. Guardrails have too often followed harm rather than preceded it. If companies want a different outcome with AI, the ordering must change. Parents are not rejecting technology. They are asking for evidence that restraint and foresight are being applied before new technologies become normalized in children’s lives. Building in that way may be slower. It may produce less dramatic growth curves. But when the users are children, speed can’t be the only measure of success.

A deeper dive

Here are a few helpful resources in case you want to really dig into today’s topic:

• New Mexico is suing Meta, claiming that the company lied about efforts to improve child safety on its platforms. State attorney Donald Migliori told the courtroom, “The theme throughout this trial is going to be that Meta puts profits over safety.” He argued that the company knowingly overestimated the impact of its teen safety features and that safety was underfunded and ignored internally. The trial is ongoing, but the accusations are old news. Meta has broken trust with parents too many times to ever be taken at their word.

• In an effort to protect children and families from the negative aspects of technology, Australia enacted a sweeping social media ban for users under 16. But Roblox, one of the most popular platforms for children, was not included in the ban. The embattled company is still drawing lots of scrutiny, as reports surface of inappropriate content, harassment and grooming. The Australian government has now requested an urgent meeting with the company to discuss these issues.

TL;DR

Too long; didn’t read. It shouldn’t be a full-time job to keep up on industry news, so here is a mercifully quick summary of some other notable developments:

• Meta, TikTok and Snap have agreed to be rated on teen safety by a coalition of advocacy organizations. The platforms will be graded on things like whether they mandate breaks and provide options to opt out of infinite scroll. Dan Reidenberg, managing director of the National Council for Suicide Prevention, oversaw the development of the standards, and said, “I hope that by having this new set of standards and ratings it does improve teens’ mental health. At the same time, I also really hope that it changes the technology companies: that it really helps shape how they design and they build and they implement their tools.” It will be interesting to see if the new rating system changes teen behavior or decision-making at the companies themselves.

• You might know MrBeast as a stunt-pulling YouTube personality, but now he’s pivoting into finance. The star has just acquired banking app Step, which helps teens open their first bank account and establish credit. And apparently, MrBeast plans to launch another YouTube channel focused entirely on youth financial literacy.

And lastly

Here are a few more pieces of original writing from me and my team—just in case you’re keen for more:

• When we asked parents how they felt about AI for kids, we heard a lot of interesting takes and passionate opinions. We’re working on the full report, but you can check our blog post here for initial information.

• And if you’re wondering who MrBeast is, have a look at our parents’ guide here.

For a while now, researchers, child-safety advocates and some parents have been warning that generative AI would make it easier to create deepfake images, including fake nudes, using ordinary photos pulled from the internet. Once image-generation tools became fast, public and easy to use, misuse was not a remote possibility. It was an eventual outcome. And that is exactly what we are seeing now.

In recent weeks, Grok has been used to generate and publicly post sexualized images of real women and children, often from minimal prompts. British regulators are now investigating whether X violated online safety laws designed to prevent the spread of nonconsensual intimate images and child sexual abuse material. Other countries, including Indonesia and Malaysia, have already banned the tool outright.

This matters because it marks a shift in how governments are responding. The question is no longer whether platforms are doing enough to remove harmful content after the fact. Regulators are asking why tools that predictably enable abuse are being offered in the first place. For parents, however, this moment raises a different and more personal concern.

AI has changed what online images mean. Photos used to function as communication or memory. Today, any image shared publicly can be scraped, altered and repurposed at scale. Deepfake technology removes the need for familiarity, access or targeting. An image alone is enough. That reality complicates a common and largely unexamined behavior: parents sharing photos of their children online.

Most families do this for ordinary reasons. Staying connected with relatives. Sharing milestones. Letting distant grandparents feel close. Those motivations have not changed. What has changed is the environment those images enter.

Once a photo is online, it may circulate far beyond its intended audience. Context collapses. Consent is assumed rather than granted. Children who cannot meaningfully agree to their digital presence bear the long-term consequences of decisions made on their behalf. But this is not about blaming parents. It is about acknowledging that the risk profile has shifted, often without clear warning or guidance.

The Grok episode illustrates how quickly and easily that shift can turn into harm at scale. When generative tools are embedded into social platforms, made public by default and optimized for engagement, abuse no longer depends on bad actors seeking out victims. Systems do the work for them.

Elon Musk helped accelerate the problem himself, publicly promoting Grok’s image-generation capabilities just as concerns about deepfakes were intensifying. What followed was not a minor misuse, but a rapid escalation. Sexualized images of real women and children circulated widely, drawing international scrutiny.

Only after that pressure mounted did Musk move to limit the feature, announcing that image generation would be restricted to paying, premium users. The change was presented as a safeguard. In reality, it treated access as the problem, rather than the underlying design of a tool that predictably enabled abuse. British officials criticized the move, noting that monetization does nothing to address the underlying harm.

The lesson here is broader than one chatbot or one company. When platforms prioritize speed, scale and novelty, the costs are often externalized onto users, and especially onto children. Parents are then asked to mitigate those risks individually, through privacy settings, vigilance and restraint, while the systems themselves remain largely unchanged. That is an unsustainable balance.

Images of children should not be treated as raw material for generative systems that can sexualize or distort them with ease. Parents should not have to weigh the risk of abuse every time they share a photo meant for people they know. And kids should not grow up with digital footprints shaped before they could understand the trade-offs involved.

Deepfakes and fake nudes are no longer fringe concerns. They are a predictable consequence of how generative AI is being deployed by unscrupulous companies. If we want a different outcome, the focus must shift upstream, toward design decisions, incentives and limits, rather than continuing to react after harm occurs.

A deeper dive

Here are a few helpful resources in case you want to really dig into today’s topic:

• The non-consensual images have flooded X (formerly Twitter) and the victims aren’t getting much response from the platform. According to The Atlantic, X owner Elon Musk seems to be egging the whole thing on, making jokes as the whole thing was unfolding. While the platform eventually took some of the images down, many, many more remain live on the site.

• You can’t help but wonder how all this is allowed to happen. Well, this is what it looks like when there’s no meaningful regulation of AI. In a strange turn, OpenAI is partnering with Common Sense Media to present new legislation in California meant to protect children from AI like Grok. Common Sense Media originally had its own bill, which included private right of action, but the new version, co-signed by OpenA,I would hold companies accountable through the AG and financial penalties.

TL;DR

Too long; didn’t read. It shouldn’t be a full-time job to keep up on industry news, so here is a mercifully quick summary of some other notable developments:

• Want to boost your cognitive ability? Video games might be the ticket. According to new research, “playing action video games in particular may prove beneficial for a wide range of skills, such as our attention for visual information and our ability to learn.”

• Meta has removed more than half a million children from its platforms in Australia, as part of an effort to comply with the new ban in that country. According to ABC News, “The social media giant says it removed 330,000 users from Instagram, 173,000 from Facebook and 39,000 from Threads between December 4 and 11.”

And lastly

Here are a few more pieces of original writing from me and my team—just in case you’re keen for more:

• AI is in the news constantly, but it’s not always easy to understand how this technology works. Here’s a simple explainer from my team to help cut through the noise.

• If you’re looking for a safer way to share pictures of your children online, here’s a guide from my team on how to minimize the risks.

Happy New Year, and thanks for being here. I appreciate everyone who’s followed along as we work to make technology safer and healthier for our families. This newsletter exists to help parents navigate the digital world with confidence. That work feels more urgent now than it did even a year ago, as the tools our kids use every day become more complex and influential.

For years, the debate about kids and technology has focused on screen time while missing what’s happening behind the screens. Digital platforms are now highly social, algorithm-driven and immersive, making it harder for families to understand how these systems shape kids’ experiences. In 2025, that shift became more pronounced. Kids are spending time in digital spaces designed to capture attention, encourage comparison and sustain engagement. Simply counting hours doesn’t reflect what kids are exposed to or how those environments influence behavior.

That’s why 2026 matters. Governments are reexamining age limits that no longer align with modern platforms. Companies are facing pressure to explain how their products actually work. AI is changing the scale and nature of online content. And, parents are frustrated by the status quo. As we head into the new year, I wanted to share my predictions for where technology (with a special focus on kid tech) goes next:

1. More countries will raise the minimum age for social platforms

The default age of 13 belongs to a much earlier internet. Today’s platforms aren’t just places to post updates. They are algorithm-driven environments designed to maximize engagement, comparison and emotional response at scale. Governments are increasingly recognizing that these dynamics create a fundamentally different risk profile for younger users than the web of the early 2000s.

In 2025, Australia became the first country to formally raise the minimum age for social media to 16. That move matters less because it is perfect policy and more because it broke a long-standing stalemate. Once one country takes the political risk, it becomes easier for others to follow. Parts of Europe, particularly Scandinavia, are well positioned to move next given their strong child-welfare norms and digital infrastructure. Canada often mirrors European momentum. In the U.S., federal change will be slower, but pressure continues to mount as parents, researchers and state governments push for stronger protections for teens. Look for one or more of these countries to build on this momentum and implement their own minimum age-adjustments for social media.

However, we need to consider that as much as these bans sound great on paper, they do come with real tradeoffs. Higher age limits can push kids toward less visible platforms, encourage workarounds or exclude them from digital spaces that have become central to social life. Without clearer definitions of age-appropriate design and better alternatives for younger users, bans alone won’t fix the underlying problems.

Still, Australia’s move forced a conversation many governments and platforms had avoided. It made clear that the status quo of pretending modern social platforms are appropriate for early teens by default is no longer tenable.

2. There will be a continued push for age verification from app stores

Raising age limits inevitably raises a second question: who is responsible for verifying age? In 2025, we saw platforms experiment with everything from self-attestation to AI-based facial analysis. The result has been a fragmented, confusing experience for families, and a compliance burden that falls hardest on smaller developers.

When every app is responsible for age checks and parental consent, enforcement becomes inconsistent and data risk multiplies. Each app collects a little bit of sensitive information, often without the scale or resources to protect it properly. Meanwhile, bad actors are rewarded for cutting corners.

Apple and Google already control more than 95% of the app distribution market. If you have an account with either company, they already know who you are, what devices belong to your family, and how parental controls are configured. Moving age verification upstream to the app stores would reduce duplication, simplify expectations, and turn parental controls into real infrastructure rather than a patchwork of settings.

I’m generally not enthusiastic about giving Big Tech more power, and centralizing age verification raises legitimate privacy concerns. But decentralization has failed to deliver either safety or simplicity. A single, regulated enforcement layer may actually better align with data minimization and usability at scale.

Apple and Google may resist this shift, and that resistance is understandable. Being accountable for age gating increases regulatory risk with little immediate upside, especially compared to the comfort of collecting a 30% cut from an app ecosystem they already dominate. But this is also the flip side of scale.

While they enjoy the advantages of a near-total app store oligopoly including distribution control, identity infrastructure, and unmatched reach, this is one of the few areas where that dominance creates not just power, but responsibility. If age verification is going to work at all, it needs to live where scale, consistency, and enforcement already exist. This is a responsibility Apple and Google are uniquely positioned to shoulder, and one they should welcome as the price of their central role in the app economy.

3. Roblox will face a category reckoning

Roblox occupies one of the most fragile positions in the kid-tech ecosystem. It is branded as a game and creative platform, but in practice it functions as a massive social network for children. Persistent identities, friend graphs, real-time chat and voice, user-generated content, and virtual economies shape the daily social lives of millions of kids. For many users, Roblox isn’t something they play, it’s where they hang out.

As governments increase scrutiny on social platforms, this ambiguity becomes harder to sustain. If algorithmic social environments are deemed too risky for younger teens, it raises an unavoidable question: what category does Roblox actually belong to? Regulation based on how a product is used, not how it’s marketed, is becoming the dominant lens. Roblox sits squarely in the crosshairs.

Leadership and public framing matter enormously in moments like this, and Roblox has not helped its own case. In a recent appearance on the Hard Fork podcast, CEO David Baszucki came across as dismissive of responsibility, repeatedly deflecting concerns back onto parents rather than engaging seriously with the platform’s social dynamics. More troubling, his comments expressing curiosity and excitement about prediction-market–style mechanics (effectively flirting with the idea of gambling-like systems for kids) landed badly and reinforced fears that Roblox underestimates the risks embedded in its own ecosystem.

Moments like that don’t create regulatory pressure on their own, but they accelerate it. When executives appear disconnected from how their products are actually experienced by children, trust erodes quickly with parents, policymakers, and investors alike. Even without immediate regulatory action, prolonged uncertainty around classification, duty of care, and future constraints is likely to weigh on Roblox’s public perception and long-term valuation.

4. AI slop will accelerate the collapse of social media as a trusted utility

Social media already feels strained. Feeds are saturated with ads, dominated by non-friend content, and increasingly optimized for engagement rather than relevance. Many users tolerate this because platforms still serve a basic utility: staying informed, being entertained, and maintaining a loose sense of shared reality.

AI-generated slop threatens that last remaining value. When images, videos, and audio can be produced endlessly, cheaply, and convincingly, social feeds stop functioning as reliable reference points. The problem isn’t just low-quality content, it’s epistemic breakdown. Once users can’t be confident that any piece of content is real, the entire feed becomes suspect.

This is happening at a uniquely dangerous moment. We’re already living through intense political polarization, declining institutional trust, and heightened emotional volatility online. AI-generated misinformation doesn’t just add noise, it acts as an accelerant. Even a small number of widely shared fake videos or manufactured events can trigger real-world consequences before fact-checking has a chance to catch up.

For kids, this breakdown is especially dangerous. Children are still learning how to assess credibility, understand intent, and separate signal from manipulation. When AI-generated content becomes indistinguishable from reality, kids don’t just consume misinformation. They internalize distorted norms, false narratives, and emotionally charged content without the cognitive tools to process it. In an environment already optimized for engagement, AI slop amplifies confusion rather than curiosity, making it harder for kids to develop healthy digital judgment.

The result will be a slow, but accelerating, erosion of utility. Social platforms will become less useful for news, less enjoyable for entertainment, and less trustworthy as shared spaces. Combined with ad overload and algorithmic prioritization of strangers over friends, I believe the rapid growth of AI slop will push feeds past a breaking point.

As trust collapses, behavior changes. People retreat. Networks tighten. Private group chats, direct messaging, and curated sources replace public feeds. Social media doesn’t disappear overnight, but it stops serving as common ground. In that sense, AI slop may mark the beginning of the end for social platforms as broadly useful, shared infrastructure. Not because people stop scrolling, but because scrolling stops meaning anything.

5. AI hype will cool as it becomes more integrated in the digital world

AI capability is still improving, but the pace of breathtaking breakthroughs is slowing. As generative tools become embedded in everyday software, AI will shift from something we actively seek out to something that quietly shapes how we write, search, create and communicate.

For most people, AI won’t feel like a destination. It will show up inside email, documents, messaging apps and creative tools, often without a clear moment of interaction. The biggest winners will be companies with massive install bases and data (think Microsoft, Google, etc.), while consumer-facing novelty fades into the background.

For families, this makes understanding AI more important, not less. Kids will encounter AI long before they think to question it. AI will show up in search results, creative prompts, homework tools and conversations. That puts new weight on helping them learn how to use these systems thoughtfully, recognize their limits, and understand when AI is influencing outcomes.

The real risk isn’t that AI becomes too powerful, but that it becomes too ordinary. As AI fades into the background of everyday tools, its influence becomes harder to see and easier to ignore. Kids won’t experience AI as a moment of choice, they’ll inherit it as default.

That makes intentional design, transparency, and age-appropriate boundaries essential. If AI is going to quietly shape how kids write, search, create, and communicate, then adults can’t afford to be passive observers. The next phase of AI won’t announce itself with breakthroughs or headlines - it will simply become the water kids are swimming in. Our responsibility is to make sure that water is safe, understandable, and worthy of their trust.

Here’s to a safe, healthy and creative 2026

The tools kids use are changing faster than the frameworks meant to guide them, and that gap is creating real friction for families, companies and policymakers alike. For parents, the goal isn’t to anticipate every policy change or master every new product. It’s to better understand the systems shaping kids’ digital lives, and to stay engaged as those systems evolve. Confidence comes less from strict control and more from clarity, shared expectations and helping kids build judgment as technology becomes a normal part of their lives.

That’s what I’ll continue focusing on here in the year ahead: making sense of what’s changing, surfacing the tradeoffs that matter and helping parents navigate kid tech with a little more confidence as the ground continues to shift.

TL;DR

Too long; didn’t read. It shouldn’t be a full-time job to keep up on industry news, so here is a mercifully quick summary of some other notable developments:

• Instagram isn’t as popular with tweens and teens as competitors like Snapchat and TikTok. But, according to a trove of internal documents, the company is working hard to change that. According to the Washington Post, “The company directed staff to boost teen-friendly influencers, adjust Instagram’s algorithm to make it easier for new teen users to find people they know and invest in paid marketing that highlighted the app’s role in helping teens connect with friends.”

• How many social media users do you think are toxic? According to a new study out of Stanford University, the number might be smaller than you think. On average, people estimated that 50% of users on Reddit had posted toxic content. The actual number? Much smaller at 3%. That’s a 13-fold overestimation, and it goes to show you how a few vocal jerks can seriously affect the vibe.

And lastly

Here are a few more pieces of original writing from me and my team—just in case you’re keen for more:

• Kids are increasingly exposed to AI, and it can feel overwhelming to know what’s safe and what’s not. My team put together this guide with the top 5 red flags to look out for.

• At Kinzoo, we believe strive to make it easier for parents to navigate technology with their kids, and that means designing platforms that keep them in the loop seamlessly. Here’s how that works.

Happy holidays to all of my readers. I hope this season gave you the chance to slow down and spend meaningful time with friends and family.

If your holiday gatherings included conversations about technology and parenting, you may have found yourselves discussing Australia’s recent decision to ban social media for children under 16. It’s a significant move, and one worth unpacking. Let’s dive in.

Australia’s ban on social media for children took effect on December 10 and has been praised as bold and long overdue. Given how anxious many parents feel about their kids’ online lives, that reaction is understandable. Platforms have repeatedly failed to earn trust, harms are well-documented and policymakers are under pressure to act. But while bans like this look strong on paper, they are often more complicated in practice.

In reality, bans tend to push kids into one of two camps. Some work around the restrictions, migrating to less regulated platforms that are harder for parents and policymakers to see. Others are excluded altogether from digital social spaces that have become a central part of modern life. Neither outcome addresses the underlying problem.

History shows that kids are remarkably good at navigating around rules. Most parents can attest to this firsthand. Age gates are easy to bypass, accounts can be shared and VPN usage often spikes when restrictions are introduced. When platforms are blocked, behavior rarely disappears. It shifts, often to spaces that are less visible and less equipped to moderate harm. Even when bans reduce exposure in the short term, they often delay rather than resolve the issue. Postponing access does little to prepare kids for the digital world they will eventually enter.

When access returns, it often does so all at once, without guidance, context or gradual skill-building. Digital literacy and social judgment are not developed through avoidance. They are learned through supported participation. Protection without preparation can leave young people more vulnerable later on.

There is also a cost to exclusion that is rarely acknowledged. Social media is not just entertainment. It is where culture circulates, friendships are maintained, identities are explored and creativity is shared. Digital social spaces have become central to how young people relate to one another. Removing kids entirely from these spaces risks isolating them from peers and from the cultural conversations shaping their generation. Acknowledging that does not mean minimizing harm. It means recognizing that exclusion carries consequences of its own.

Notably, this blunt approach was not inevitable. Reporting from Cam Wilson shows that earlier versions of the legislation included an exemption framework that would have allowed platforms to avoid the ban if they met specific safety and design standards for children. That approach could have incentivized the creation of genuinely kid-appropriate social spaces rather than imposing a universal cutoff. The exemption was removed before the bill was finalized, leaving prohibition as the primary tool.

The limits of this approach become clearer when you look at what Australia’s ban does not include. Platforms like Roblox and Discord are excluded, despite being among the most common places kids actually socialize. Both function as large-scale social environments and have faced well-documented safety challenges, including grooming, scams, and exposure to inappropriate content.

The issue is not the individual platforms themselves. Harm follows social interaction at scale, especially when platforms are designed to maximize engagement and operate with uneven moderation. Drawing regulatory lines around certain apps while excluding others with similar dynamics exposes the limits of platform-by-platform solutions.

At the center of this debate is a problem that bans don’t fix: platform design and incentives. Most social platforms are built to increase time spent, frequency of interaction and emotional response, because those metrics drive revenue and growth. When engagement is the goal, design decisions drift toward whatever keeps users coming back, regardless of the downstream effects on wellbeing.

As artificial intelligence becomes more embedded in social systems, this dynamic intensifies. AI makes platforms more persuasive, more personalized, and more adaptive. In that environment, banning a short list of apps grows less effective. The risks move with design patterns, not brand names.

If we want to meaningfully reduce harm for kids, the answer lies in platforms built with child-centered defaults, limits on engagement-driven optimization and safety embedded into the core experience. Australia’s ban signals that something is wrong. Real progress will come from demanding platforms that do less harm, rather than simply reducing the number of places kids can go.

A deeper dive

Here are a few helpful resources in case you want to really dig into today’s topic:

• Most parents won’t be surprised to learn that a lot of teens in Australia don’t plan to comply with the ban. They’ve been brainstorming ideas about how to circumvent it, including alternate apps, fake logins and even signing into their parents’ accounts. As one 15-year-old explains, “Teens aren’t trying to rebel for no reason. We just want to stay in contact with our friends and exist in the world the way every generation before us did. A ban won’t fix the issues they think it will.”

• While the law has broad support in the Australian parliament, as well as from parents, tech companies aren’t as thrilled with it. Reddit is suing the government already, arguing that the ban interferes with protected speech. They also argue that they don’t meet the criteria to be included in the ban. I can’t help but wonder why Reddit is included in the ban, while Discord and Roblox managed to squeak by.

TL;DR

Too long; didn’t read. It shouldn’t be a full-time job to keep up on industry news, so here is a mercifully quick summary of some other notable developments:

• The Wizz app has been described as “Tinder for kids.” It’s basically the same premise as Tinder (aka browsing strangers’ profiles in order to connect), but unlike Tinder, Wizz is open to teens. Not surprisingly, there have been several cases of children meeting up with people they thought were other children, only to be assaulted by adults. It doesn’t take a rocket scientist to see how this kind of app could be so dangerous for children. They just shouldn’t be connecting with a bunch of strangers online.

• Children in China are using bots and engagement hacks to get more likes on their smartwatch profiles. “On the extreme end, Chinese media outlets have reported on kids who buy bots to juice their numbers, hack the watches to dox their enemies and sometimes even find romantic partners.” While smartwatches can be a great tool to help parents stay in touch with kids, any platform with public like counts creates pressure and incentives that kids might not be ready for.

And lastly

Here are a few more pieces of original writing from me and my team—just in case you’re keen for more:

• The Nintendo Switch 2 promises to be a popular gift this holiday season. If you’re setting up a new device for your family, check out this parent’s guide here.

• And if you’re considering getting your kids a smartwatch this year, we have all the pros and cons to consider here.

That conversation matters even more right now. Several states have filed lawsuits accusing Roblox of failing to protect children from predation and exploitation. Kentucky’s attorney general recently announced a case focused on suspected pedophiles operating on the platform. Louisiana and Texas have sued as well, alleging that Roblox enabled systemic abuse and put profits ahead of safety. Florida has issued criminal subpoenas to assess whether the platform is aiding predators. And just last month, activist groups held a virtual protest inside Roblox to deliver a petition calling on the company to put child safety before engagement metrics. Parents are watching closely. Regulators are watching too.

I listened to the recent Hard Fork interview with the company’s CEO the same way many parents and tech folks did. At first, with curiosity. Then with surprise. And eventually with concern. It feels worth talking about, because it offered a rare look at how Roblox is navigating children’s safety, harm and responsibility at a moment when it’s under a national spotlight.

I have met David Baszucki, the CEO. He struck me as genuine and deeply committed to what he has built. I believe he cares a great deal about Roblox and about the kids who use it. At the same time, leaders in his position face enormous pressure, and that pressure can create a kind of cognitive dissonance. You can care deeply and still struggle to talk openly about harm. You can want to protect kids and still feel pulled toward growth, innovation and investor expectations. That is why public moments like this podcast are so revealing. They show instincts that do not always make it into polished statements or press releases.

There were parts of the interview that reminded me of what teams go through in the early days of building a product for kids. David talked about the founders taking turns as moderators. He described a moment when they paused features to build safety tools, even though the platform was tiny. Anyone who has ever built for kids knows how unusual that choice is and how much it speaks to the intentions at the foundation of the company. He also talked about the scale of Roblox today. A platform that now sees billions of hours of monthly activity is obviously a lot more complex. The technical work needed to filter content and prevent harm is relentless. Many people underestimate that. I understand why he wanted to emphasize it.

Still, the tone of the conversation shifted as the interview went on. Some answers sounded rushed. Some felt defensive. And there was a moment when David refused to acknowledge that predators operate on the platform, despite years of reporting and lived experience from families. Hearing that kind of dismissal was jarring. Against the backdrop of lawsuits, subpoenas and parent protests, that response stood out even more. When a platform holds the attention of children as young as five, the tenor of the public conversation matters. As someone who builds kid-tech, I do not expect perfect solutions from a platform like Roblox. But I think parents rightly expect steadiness and a willingness to acknowledge real harm. They want to hear about what is being done and why. A platform that has become a central part of childhood needs to hold itself to a pretty high standard.

Near the end of the interview, the hosts asked David a question about prediction markets, which are essentially betting systems wrapped in forecasting language. The idea of building a child-friendly version inside Roblox came up, and David seemed to love it. His excitement was real and immediate. It was also revealing. When the first reaction to a gambling-adjacent concept for kids is enthusiasm, you know that the pull of engagement is winning out over safety. Kids respond intensely to reward cycles. They form habits quickly. Any feature or game that taps into that deserves a long pause at the very least. David’s reaction to this hypothetical showed me where the centre of gravity has shifted. And perhaps it’s time for a correction on that.

I believe Roblox has people who work hard every day to make the platform safer. I also believe the company would benefit from approaching public conversations about harm with a deeper sense of understanding and empathy. Children and families have experienced real consequences across many platforms over the past decade. And taking that seriously helps restore trust. And it shows that you understand how important and consequential this work can be.

When you build tech for kids, you need to be able to sit with the hard questions. You need to stay patient, stay open and stay grounded when safety comes up. I believe that, when you build technology for children, the way we talk about that work is part of the work. If platforms can meet these moments with care and clarity, families will feel it. And that feeling is the foundation of trust.

A deeper dive

Here are a few helpful resources in case you want to really dig into today’s topic:

• A transcript of the interview is here, along with a video of the exchange. If you have the time, it’s worth a listen, because it gets beyond the highlight moments and paints a picture of how complex the kid-tech space can be.

• Some people have decided to take Roblox’s predator problem into their own hands. Frustrated by what they think is a lack of action on the company’s part, a few folks have started “hunting” predators on the platform by posing as young children and even arranging meetings in real life to capture these alleged groomers on camera. Not surprisingly, Roblox doesn’t love this. They’ve banned one of the most famous hunters for violating their terms of service.

TL;DR

Too long; didn’t read. It shouldn’t be a full-time job to keep up on industry news, so here is a mercifully quick summary of some other notable developments:

• When companies focus on engagement over safety, it creates platforms that are unsafe for children and families. Case in point: Meta and its chatbots. The company has a pattern of making decisions that can be harmful to kids, and then attempting to suppress research about them. Recall that they initially allowed their chatbots to have romantic and sexual conversations with kids, and that’s a decision I can’t believe made it all the way into a live product.

• A new study shows a correlation between early smartphone ownership and higher rates of depression, poor sleep and obesity among tweens. The research indicated that the younger the children were when they got their first smartphones, the greater their risk of obesity and poor sleep. The correlation is likely because screen time is replacing other activities like physical exercise and sleep. Dr. Ran Barzilay, the lead author for the study, says, “When you give your kid a phone, you need to think of it as something that is significant for the kid’s health—and behave accordingly.” When it comes to smartphones, parents need to be mindful of how they’re used and what other activities they might be edging out.

And lastly

Here are a few more pieces of original writing from me and my team—just in case you’re keen for more:

• The holiday season is here, so lots of us are probably trying to figure out what to get our kids. My team put together this gift guide, with suggestions broken down by age and safety feature.

• Winter break usually involves a lot of slow, cozy days, and that often comes with an uptick in screen time. Before you stress, check out this guide.

We’ve all been there: scrolling the app store, skimming ratings, squinting at “ages 9+,” and wondering if any of it means… anything at all. Every tool promises to do something great for kids, but the truth is simple: most apps aren’t designed with families in mind. They’re engineered for engagement. Their job is to keep your child’s attention. Your job is to decide whether something earns a place in their world.

Parents are told we’re “in control,” but the moment usually comes after the download, and after your child is already tapping through sign-up screens, swiping past prompts and navigating design patterns built to pull them deeper. You’re left trying to manage screen time, block content and stay ahead of algorithms built to learn faster than any human can parent.

Meanwhile, your goals are completely different from the people designing those systems. You’re asking: Does this fit our family? Does it support the kind of relationship I want my kids to have with technology? Will it help them grow or just keep them busy?

Those questions don’t require a framework or a checklist. You just need a pause, a breath and a bit of clarity. A few guiding ideas help:

1. Follow the incentives.

If you’re not paying with money, you’re paying with attention or data. And when it’s your child’s attention or data, the trade-off deserves a closer look. Who benefits when your kid opens this app, and how do they measure success?

2. Watch what it’s teaching.

Every tool claims to do something: educate, entertain, connect. That’s important, but you also want to look at how your kid will interact with the technology. Look at the behaviors the app rewards. Look at the loops it builds. Those loops become habits.

3. Use the settings that actually shape the experience.

This is where the distinction matters. There are two layers of parental controls, and most parents only ever see the second one.

In-app controls manage what happens after your child is already inside a product, with things like chat settings, friend requests, content filters and privacy toggles. They’re important, but reactive. They help you steer the ship once you’re already in the water.

Device-level and app-store-level controls do something more powerful: they shape the entire environment before anything reaches your child. You can block downloads above a certain age rating. Require approval for anything new. Disable in-app purchases across the board. Remove browser access entirely. These settings are your first line of defence. They give you room to think before your child taps “get.”

The strongest, simplest protections live at this system level. They’re built into your phone, your tablet or your family account, and they complement the parental controls inside any app you consciously choose to download.

I’m always surprised by how few parents are taking full advantage of these settings, but I think it’s a visibility issue. If you do one thing this week, open those settings. Review and adjust them. Set them up, not as surveillance or punishment, but as boundaries that make room for healthy choices. And talk to your kids about it. Kids do better when they understand the “why.”

You don’t need a tech background to make good decisions about tech. You just need to remember that your values outrank anyone’s engagement goals. Every limit you set teaches your child something about agency, safety and self-respect. Every conversation you have reminds them that your family gets to decide what belongs in their digital world.

Tech companies build for attention. Parents build for growth. The more we slow down, use the tools we already have and make intentional choices about what comes in the door, the more we shape a digital world that works for families on our terms.

A deeper dive

Here are a few helpful resources in case you want to really dig into today’s topic:

• I suspect a lot of parents think that device and app store parental controls are onerous to set up. They’re actually easier to set up than you’d think. You can find instructions for App Store controls here, and Android devices here.

• One of the most pressing tech issues for parents these days is undoubtedly AI, and almost all of the mainstream platforms were launched without any kind of parental control. And now they’re playing catch-up. ChatGPT is launching tools to help parents set time and content limits on their teens’ accounts and receive notifications if there are signs of potential self-harm. When did they bring these online? After grieving parents filed a wrongful death suit alleging that the chatbot helped their son plan his suicide.

TL;DR

Too long; didn’t read. It shouldn’t be a full-time job to keep up on industry news, so here is a mercifully quick summary of some other notable developments:

• It’s no secret that kids are turning to AI for mental health support, and a new study from Common Sense Media and Stanford Medicine’s Brainstorm Lab for Mental Health Innovation tested how well the chatbots do in those scenarios. The results? Not great. In fact, researchers said it’s one of the most dangerous uses of technology. The study found that “AI chatbots struggled to appropriately address mental health because they appear to be designed to maximize engagement, not safety.” That checks out. Those platforms weren’t made to be healthy; they were made to keep us chatting, no matter what.

• Most parents will tell you that screen time is one of the most contentious topics in their household. Whether you’re enforcing time limits or battling over a new app they want to download, you’ve probably witnessed a tech-induced meltdown more often than you’d like. That’s why some parents are opting for professional screen time coaching or digital detox camps. One reason it feels so tough to keep on top of tech is that the platforms aren’t designed for kids. As parents, we scramble to understand and block new technologies. But when tech is designed from the ground up for families, it makes it easier and more intuitive for us to manage it for our kids.

And lastly

Here are a few more pieces of original writing from me and my team—just in case you’re keen for more:

• If you’re curious about how AI is showing up in the classroom, my team put together this parents’ guide here.

• Parental control apps aren’t a magical solution to your screen time problems, but they can be an invaluable tool to keep kids safer. If you’re considering one, this article is for you.

That finding comes from a new study by the Center for Democracy and Technology, and while the sample size isn’t enormous, the story it tells is unmistakable. Kids aren’t just using AI to finish essays or brainstorm ideas. They’re turning to it for comfort, connection and company. And in a world where loneliness is already an epidemic, that should give us pause.

For a lot of kids, AI feels safe. It listens. It responds instantly. It never rolls its eyes, never judges, never says the wrong thing. If you’re fourteen and feeling unseen, that kind of reliability can be magnetic. But AI doesn’t care about them. It’s been carefully trained to sound like it does. It mirrors empathy without feeling it, offering the illusion of intimacy without the risk or reciprocity that make real relationships real. All this means kids are learning to connect with something that doesn’t actually know what connection is.

The CDT report also found an interesting correlation: the more schools use AI, the more likely students are to say they know someone who considers AI a friend or romantic partner. Teachers use it to grade papers and create lessons. Students use it to draft assignments and study. Over time, it becomes normal to chat with chatbots. Yet only about eleven percent of teachers have been trained on what to do if a student’s use of AI starts to affect their wellbeing. Essentially, we’ve given kids a deeply personal tool and almost no guidance on how to use it safely. The schools that are embracing it aren’t necessarily teaching kids how to use AI. And the AI systems they’re using? They’re not always age-appropriate.

And now, the consequences are starting to show. These chatbots aren’t built to nurture, they’re built to engage. Every response is optimized to keep the conversation going, to make users feel seen just enough to stay. For lonely kids, that can spiral fast. They start confiding in a system that never says no, never logs off, never gets tired. What begins as comfort can become dependence.

We’ve already seen where that leads. One 14-year-old boy in Florida died by suicide after months of conversations with a chatbot he believed he’d join in the afterlife. His parents are now suing Character.AI, the company behind the bot.

And after months of mounting scrutiny, Character.AI announced that it will ban users under 18 from talking to chatbots altogether. It was a reactive decision, and when the announcement was made, CEO Karandeep Anand told the Financial Times, “The long-term effects of prolonged usage of AI are not understood well enough.”

The very people who built these companions are starting to admit what parents have sensed all along: they don’t actually know what happens when kids form emotional bonds with something that can’t feel anything back.

Character.AI’s move makes it the first major company to cut off chatbots for teens. But it’s less an act of moral clarity and more an act of damage control. Their hand was forced by lawsuits, regulators and grieving parents. And it underscores just how far things have gone: they’re only just restricting access to kids now.

The truth is, AI isn’t going away. It’s already part of how kids learn, create and communicate. But the line between using AI and being shaped by it is getting thinner by the day. And while companies scramble to measure the fallout, parents are left trying to hold the boundary between curiosity and harm.

A deeper dive

Here are a few helpful resources in case you want to really dig into today’s topic:

• It’s not just Character.ai facing scrutiny and lawsuits. ChatGPT has also come under fire after the platform discussed suicide at length with a teenager who later took his own life. According to their own internal research, a staggering [1 million users discuss suicide](https://www.ft.com/content/64065ad7-2f41-44b2-b58c-f5c122f5b923?) with the chatbot every week. The company said it’s trained its model to “better recognise distress, de-escalate conversations, and guide people towards professional care when appropriate.” To me, this seems like a high-stakes experiment to be running with their 800 million weekly active users.

• If you’re interested in the full report on school’s embrace of AI, it’s here. The summary says that “the frequency and variety of AI uses continues to grow; at the same time, the increased use of AI in educational settings is correlated with heightened risks to students.” At this point, we only have correlation, not causation, but this reminds me an awful lot of the rhetoric around social media harms to children. And I believe that, where there’s smoke, there’s fire.

TL;DR

Too long; didn’t read. It shouldn’t be a full-time job to keep up on industry news, so here is a mercifully quick summary of some other notable developments:

• Instagram has made another change aimed to protect teenagers using the app: they’re going to begin limiting the content younger users can see to PG-13 only. This apparently also applies to the conversations they can have with the platform’s AI chatbots. I hope that Meta will follow suit with Character.ai and restrict access to the bots altogether.

• Did you know there’s no federal law around likeness? That could be a huge problem, given that OpenAI has launched an app that can generate convincing deepfakes of you in seconds. In the absence of any law around that kind of thing, “OpenAI and other companies are responding by implementing their own likeness policies,” which is troubling, because I’m not sure I want those companies setting the standard for what is and is not appropriate when it comes to my face. Or my kids’ faces.

And lastly

Here are a few more pieces of original writing from me and my team—just in case you’re keen for more:

• Ever feel like you have no idea who your kids’ celebrities are? It’s a common feeling, so we’re putting together profiles on some of the big ones. First up: MrBeast.

• This November, the last season of Stranger Things is dropping. If you’re wondering whether to watch with your kids, here’s a parent’s guide with a breakdown by age.

The latest Pew Research Center report, How Parents Manage Screen Time for Kids, maps that reality in numbers. Screens start younger, AI is already part of childhood and most parents say they’re doing their best while also wishing technology made that easier. But just in case reading research reports isn’t your idea of a good time, I’ve put together the key takeaways below.

The on-ramp needs to start even earlier

Even among the youngest children, digital life starts early. Pew found that 82% of parents of kids under two say their child watches TV, and about four in ten say their baby or toddler already uses or interacts with a smartphone. YouTube use among that same group has jumped from 45% in 2020 to 62% today.

This is interesting, given that parents are often told “no screens under 2.” That’s the expert guidance. But the fact that kids under 2 get screen time isn’t a parenting failure, it’s a reflection of modern family life. Work, weather and convenience all play a role. What it really shows is that we need to treat digital literacy as something that begins right away. We need safe, guided experience to start early.

Digital literacy expert Diana Graber calls this the “digital on-ramp,” the early and intentional introduction to technology that helps kids learn to use it thoughtfully. It’s where children first merge onto the road of connected life, and the goal isn’t to keep them off it, but to make sure they enter safely, at an appropriate speed, and with adults alongside them. A healthy on-ramp isn’t about restriction, it’s about modeling curiosity, creativity and balance from the very beginning.

When a child’s first experiences with tech are creative, positive and connected to family, they build habits that can make the rest of their digital journey safer and more intentional.

The AI shift

The same report found that about one in ten parents of 5–12-year-olds say their child uses AI chatbots like ChatGPT or Gemini.

That number might sound small now, but it’s significant. AI is no longer abstract, it’s embedded in how kids search, create and learn. But again, the challenge isn’t stopping them from using it; it’s making sure the experiences they have are age-appropriate and well designed.

That’s why we built Kai, Kinzoo’s creative AI art tool, as a safe and moderated environment. It gives kids a way to explore generative technology without the adult internet attached. The goal isn’t to keep kids away from innovation. It’s to make sure their first encounters with it are creative, empowering and safe.

The guilt gap

Pew’s data captures something every parent has felt: 42% say they could do a better job managing their child’s screen time, while 58% say they’re doing the best they can.

That tension between “doing enough” and “doing better” is less about discipline than it is about design. The platforms most kids use are engineered to keep them watching, playing or scrolling. They don’t have natural stopping points; they have engagement loops.

So when parents feel like they’ve lost control, it’s often not because they’re inattentive. It’s because they’re trying to apply rules to digital systems that were never built for balance. This is where we need to shift the narrative: parents aren’t failing; they’re navigating environments that fail to support them.

YouTube, YouTube everywhere

Of all the platforms in the report, YouTube stands out. Pew found that 85% of parents say their child watches YouTube, including 51% who say that happens daily. Even among kids under two, six in ten watch.

For many families, YouTube is the default. It’s familiar, convenient and seemingly infinite. But that infinity is the problem. Its recommendation engine isn’t built around age or appropriateness; it’s built around retention. Educational clips can quickly lead to content that’s commercialized, overstimulating or simply not meant for young audiences.

That’s why it’s worth looking for digital spaces built with children in mind: places where content is curated, ads are limited or absent, and parents can stay in the loop. Kids deserve experiences designed for their curiosity and creativity, not their attention span.

What parents want next

Here’s where the data turns hopeful. Two-thirds of parents (67%) say tech companies should do more to set rules around what kids can see and do online, and 55% say lawmakers should step in as well.

That’s not a cry for less screen time, it’s a call for better technology. Parents are no longer content to manage everything themselves. They want systems that are trustworthy by design, not just by filter. We’re seeing the beginning of a cultural shift from parents trying to police screen time to demanding accountability from the companies that shape it.

From policing to partnerships

For years, the conversation around kids and screens has been about restriction: how long, how often, how much. But the new data suggests it’s time to change the frame.

The goal isn’t to get kids off screens; it’s to make screens worth their time. When technology respects a child’s stage of development and a parent’s need for safety, everyone wins.

That’s the idea behind what we call lifeguard parenting: letting kids explore age-appropriate digital spaces while sticking close enough to jump in when needed. With the right tools and the right philosophy, technology doesn’t have to be the enemy. They can be part of a healthier digital childhood—one designed with families, not engagement metrics, in mind.

A deeper dive

Here are a few helpful resources in case you want to really dig into today’s topic:

• If you’re interested in scanning the full report, you can check it out here. The numbers are certainly interesting, but I always find the quotes from parents give great insight. Lots of parents feel the tug of war, knowing that tech can be both dangerous and beneficial, depending on how it’s used. One parent quoted in the report says “[My daughter] learned to read through one of her apps that I let her download because she has ADHD, and her learning is just different.” The report notes that the same parent also worried when “they’re so zoned in.” The tug of war is real.

• One thing the report makes clear: kids are watching a lot of YouTube. And guess what? That platform is apparently flooded with AI-generated videos. According to Titania Jordan, Chief Parenting Officer at Bark Technologies, “AI-generated videos aimed at kids are becoming increasingly common, especially on platforms like YouTube, TikTok, and Instagram, where content is often auto-generated to grab attention or maximize views.” It’s always a good idea to co-view with younger kids, or have them watch nearby without headphones, so you can check out the content they’re consuming online.

TL;DR

Too long; didn’t read. It shouldn’t be a full-time job to keep up on industry news, so here is a mercifully quick summary of some other notable developments:

• As parents, we just want our kids to grow up happy and healthy. And if they wind up with a nice, secure, well-paying career, well that’s just great. The problem is that it’s getting harder and harder to give kids career advice, and that’s because AI is upending a lot of industries. Things that used to be sure bets, like computer science, aren’t as secure as they used to be. And, experts are suggesting students train up in the liberal arts, betting that future employers will value critical thinking skills.

• I recently wrote about Sora, OpenAI’s new social platform. It turns out the company has angered the Japanese government, which asked that it stop allowing copyrighted Japanese artwork to appear in the app. Initially, OpenAI said that companies could opt out of having their intellectual property in the app, but they’ve flip-flopped on that strategy, and it’s unclear how this thorny issue will play out. What’s clear, though, is that they launched this product without a roadmap for what would happen when people inevitably generate videos of copyrighted characters cooking meth or committing various other crimes.

And lastly

Here are a few more pieces of original writing from me and my team—just in case you’re keen for more:

• If your kids have been begging to watch K Pop Demon Hunters, you’re not alone. It’s a fun animated movie, but with a few intense scenes that might spook younger kids. Here’s a breakdown by age.

• And if your kids are suddenly more obsessed with Roblox than ever, there’s a reason for that, too. Here’s a parent’s guide to Halloween on Roblox, and why it’s such a big deal.

That’s it. A simple demo. A few taps. And suddenly the photo is transformed into something that never happened.

Not long ago, I wrote about this exact scenario as a terrifying future possibility: the moment kids could generate convincing fake images of each other instantly, without specialized tools or technical skill. That moment is here now.

Snapchat is positioning Imagine Lens as a new kind of creative play. But the reality is stark: generative AI is now embedded directly into one of the most-used apps by kids and teens. With this tool, anyone can take a real image and instantly turn it into something that never happened. Today it’s a photo of classmates kissing. Tomorrow it could be almost anything a bully decides to create: a fake fight, an embarrassing scenario, a humiliating moment. The danger isn’t that everyone will mistake these images for truth, (though that certainly is a concern). The issues is that bullying now has a bespoke visual tool, creating tangible and easily shareable ammunition for cruelty.

At the same time, OpenAI is testing Sora 2, its next-generation video model. For now, it’s invite-only. But early testers say the interface looks almost identical to TikTok: short-form, swipeable, endlessly addictive. The only difference? Every single video is AI-generated. And there’s another layer: users can upload their likeness and choose who gets to generate “cameo” videos with it: only themselves, mutual followers or anyone. Imagine a teenager casually toggling the wrong setting, and suddenly classmates have the power to make fake videos of them. Or worse, strangers. As Geoffrey A. Fowler recently wrote in The Washington Post, “Until recently, using AI to fake someone’s likeness was called a deepfake and considered taboo. Now OpenAI has rebranded it as a cameo and made it the social activity of the week.”

The parallels to Snapchat’s Imagine Lens are obvious. Both normalize synthetic media as part of everyday social life. And when anyone can generate fake images or videos of anyone else, social platforms stop working. Trust collapses. Every picture, every clip, becomes suspect.

For kids, this isn’t an abstract threat. It’s immediate. A single convincing fake photo can humiliate a classmate, ruin a reputation and spread through a school before the victim has any chance to respond. Tech companies keep dressing these tools up as creativity and self-expression. What they don’t say is that they’re handing kids the power to destroy each other’s reputations with almost no friction, and embedding it in the apps they use every day.

I’ve been warning for years that the easy availability of generative AI would render social media unusable. With Imagine Lens and Sora 2, we’re standing on that threshold. If companies insist on building these tools, then at the very least, they must not embed them into the social platforms where kids live. Generative AI for play belongs in walled-off sandboxes with real safeguards, not in apps that already struggle with bullying, harassment and misinformation.

Parents need to understand what’s being put in their kids’ hands. Policymakers need to wake up to the speed of deployment. And tech companies need to stop treating kids as beta testers for their most dangerous experiments. Because once trust in our digital lives is gone, it doesn’t come back.

A deeper dive

Here are a few helpful resources in case you want to really dig into today’s topic:

• We already know that Snapchat is one of the most popular social platforms for teens. This is where a lot of kids hang out and stay in touch with friends. If I were running Snapchat, I would think long and hard about releasing a text-to-image generator in an app where messages disappear… and then I wouldn’t do it. Bullying is already an issue on Snapchat, and this one feature has the potential to supercharge that. Whatever the engagement tradeoffs for the company, it’s just not worth the risks.

• Washington Post tech columnist Geoffrey A. Fowler tried out Sora 2, giving a handful of his friends permission to use his likeness to create videos. Predictably, some of them made him uncomfortable, and he was left questioning why users aren’t allowed to approve videos of themselves before they’re posted. Here’s what he discovered: “In a statement, OpenAI told me “you are in control of your likeness end-to-end with Sora.” When I asked why users don’t approve videos before they’re created, the company thanked me for the feedback. Translation: We know this is a problem. We shipped it anyway.”

TL;DR

Too long; didn’t read. It shouldn’t be a full-time job to keep up on industry news, so here is a mercifully quick summary of some other notable developments:

• After companies unleashed generative AI on us, schools struggled to come up with coherent plans. Some tried to ban the technology outright, while others attempted to incorporate the tools into the curriculum—with varying degrees of success. But one school actually asked its students to draft their AI policy. That’s a pretty forward-thinking approach if you ask me. Our whole business at Kinzoo is built on the fact that kids deserve access, agency and safety from their tech, and I like the idea of asking them what their priorities are when it comes to AI.

• I’ve written before about Elon Musk’s AI chatbot Grok. While a lot of other chatbot companies have tried to prevent users from having erotic conversations with bots, Musk seems to be leaning into that with companions that are explicitly sexual. That might seem like a strange strategy, but as Camille Carlton, policy director at the Center for Humane Technology, explains to the New York Times, “It’s all tied to the fundamental race to intimacy that we’re seeing in the A.I. industry. These companies know that emotional attachment means more engagement and more market share.” Yikes.

And lastly

Here are a few more pieces of original writing from me and my team—just in case you’re keen for more:

• If you’re interested in learning more about AI in the classroom, my team put together this guide here with everything parents need to know.

• And, if you’re curious about general Snapchat safety for kids, here’s a rundown with all the details.

Now we’ve entered a new frontier: AI that doesn’t just recommend content but pretends to be a friend. Chatbots that simulate intimacy. Creative tools that blur the line between imagination and automation. In the past, kids’ first steps online were shaped by risks like unmoderated spaces, peer conflicts and algorithms curating their feeds. Now their earliest experiences are just as likely to be with AI. And, these experiences are laying the foundation for how kids will come to understand relationships, creativity and boundaries in the digital world.

That’s why the idea of the digital on-ramp—a concept coined by Diana Graber - has never been more important. Graber describes it as a safe, structured path where kids can practice digital habits before they’re thrown into the chaos of the open internet. The on-ramp is where you get up to speed. You don’t just join at 80 miles an hour, you gradually work up to it. Introducing kids to digital tools is the same thing. We don’t drop them behind the wheel at speed, we start them slowly and help them gear up to the faster pace of things happening online.

The on-ramp matters because early digital experiences shape norms. If the first lesson kids learn online is that “friends” can vanish overnight or that attention comes from likes, streaks and algorithms, those habits will stick. If the first lesson is that an AI chatbot will always listen, always flatter and never say no, that shapes their expectations of real relationships, too.

Closed, guided spaces offer a different path. They provide boundaries, co-use opportunities with parents and guardrails that help kids practice creativity and communication safely. An on-ramp might look like a moderated messaging app that only allows communication with trusted family and friends. It might look like a creative tool with clear limits, where kids can experiment and share without being exposed to strangers. It might look like a classroom platform where exploration happens under the guidance of teachers and parents.

These kinds of experiences give kids room to practice - to make mistakes, learn and grow - in spaces where the stakes aren’t so high. They set expectations that digital life can be supportive, fun and connected, not overwhelming or unsafe.

And as AI becomes part of the picture, the need for these on-ramps is even more urgent. Without guardrails, kids risk forming their very first digital habits with chatbots that pretend to be friends, or with open-ended tools that automate creativity instead of nurturing it. That’s why AI on-ramps are desperately needed: to show kids that technology should be a partner in learning and imagination, not a replacement for human connection. That’s the philosophy behind how we designed Kai, the kid-safe AI in Kinzoo Messenger: fixed boundaries, parent-in-the-loop and creativity at the center. It’s not just another AI bot mindlessly stuffed into another tech product. Instead, it’s an on-ramp where kids can safely interact with a new technology.

At the end of the day, the lessons kids learn on the on-ramp matter more than the tech itself. Here are the three that count:

• Digital literacy before exposure: kids should learn to think critically about new technologies, including AI, in a space that doesn’t overwhelm them.

• Safety and boundaries before openness: clear guardrails help kids understand limits before encountering the risks of open platforms or unsupervised AI tools.

• Real relationships before artificial ones: kids need to build trust and connection with people who love them, not with algorithms or AI pretending to care.

The habits formed on the on-ramp are the ones kids will carry for years to come. We need to be intentional about those first steps. Industry leaders need to stop defaulting to “open by design” and start building true on-ramps. Because if we don’t, kids won’t be learning to drive safely - they’ll be shoved straight into freeway traffic with no preparation. And that’s a risk none of us should be willing to take.

A deeper dive

Here are a few helpful resources in case you want to really dig into today’s topic:

• It seems like every week, we get reminders of the stakes for kids’ online safety. I’ve shared too many stories in this newsletter showing that this is a life-or-death issue. And, it seems like AI chatbots are accelerating the danger, putting dangerous digital products in our kids’ hands. The FTC has even launched an inquiry into these platforms, “to learn how these companies are evaluating the safety and monetization of chatbot companions, how they try to limit negative impacts on children and teens, and if parents are made aware of potential risks.” I don’t think most parents need an FTC investigation to see that chat companions come with serious risks.

• Diana Graber’s work explores many of the important issues parents and kids face with technology. She’s been a leading voice in digital literacy for more than 15 years, with articles, curricula and a book to show for it. In this article, she reflects on the lessons she’s learned over the years, and I thought this was particularly poignant: “Digital literacy today is about understanding AI and using it ethically as a tool, not a toy or a companion.”

TL;DR

Too long; didn’t read. It shouldn’t be a full-time job to keep up on industry news, so here is a mercifully quick summary of some other notable developments:

• In a new report, Common Sense Media called Google Gemini “high risk” for kids and teens. While the platform did remind users that it’s a computer, not a friend, it could still share inappropriate content about sex, drugs and alcohol with younger users. And, it can share inappropriate mental health advice. According to the non-profit, kid-safe AI products need to be built from the ground up. I couldn’t agree more.

• Lately, the conversation around digital safety for kids has focused on chat companions, but there are still other dangers to contend with. One grieving mother is suing the social platforms she says fed her son dangerous subway surfing content that ultimately led to his death. Apparently, the teen started watching urban exploring videos during the pandemic, and was quickly led to videos of people riding on the outside of subway cars—an extremely dangerous activity. He began posting his own videos, and ultimately died while subway surfing. While the social platforms maintain that these videos are removed by moderators, we all know that the algorithm loves risky, alarming content, often pushing it to kids and teens.

And lastly

Here are a few more pieces of original writing from me and my team—just in case you’re keen for more:

• If you have questions about chat companions and AI safety, my team put together this parents’ guide here.

• One challenge for parents trying to keep on top of their kids’ tech use? The platforms make it difficult or impossible. That’s why we design not just for child safety, but to make it easy for parents to stay in the loop. Here’s why that’s important.